From the a Stockholm South Hospital, Stockholm, Sweden,

Transcription

1 Comparison of very early treatment with either fibrinolysis or percutaneous coronary intervention facilitated with abciximab with respect to ST recovery and infarct-related artery epicardial flow in patients with acute ST-segment elevation myocardial infarction: The Swedish Early Decision (SWEDES) reperfusion trial Leif Svensson, MD, PhD, a Mikael Aasa, a Mikael Dellborg, c C. Michael Gibson, Ajay Kirtane, d Johan Herlitz, b Ake Ohlsson, a,b Thomas Karlsson, b and Lars Grip b Stockholm, and Gothenburg, Sweden; and Boston, MA Background Results from a number of studies indicate that primary percutaneous coronary intervention (PCI) is superior to fibrinolysis for treatment of acute ST-elevation myocardial infarction. Modern adjunctive antithrombotic treatment with systematic use of low-molecular-weight heparins, fibrin-specific thrombolysis, and glycoprotein IIb/IIIa receptor inhibitors may improve the outcome compared with what was achieved in previous studies. Methods Patients with ST-elevation myocardial infarction were randomized to receive enoxaparin followed by reteplase (group A; n = 104) or enoxaparin followed by abciximab and transfer to invasive center for optional PCI (group B; n = 101). Primary end points were ST-segment resolution 120 minutes and TIMI flow at coronary angiography 5 to 7 days after randomization. Results Forty-two percent of the patients started therapy in the prehospital phase. Time from symptom to treatment was 114 minutes in group A and 202 minutes in group B. Baseline characteristics were similar in the 2 groups. Sixty-four percent in group A and 68% in group B had ST resolution of N50% at 120 minutes (not significant). At control angiography, 54% in the fibrinolytic group and 71% in the invasive group had TIMI 3 flow ( P =.04). At 30 days, the composite of death, stroke, or reinfarction occurred in 8% in the fibrinolytic group compared with 3% in the invasive group (not significant). Conclusions Despite much shorter time delay to start of fibrinolysis than PCI, this did not result in signs of superior myocardial reperfusion. Epicardial flow in the infarct-related artery was better after invasive therapy, and there was a trend toward better clinical outcome after this treatment compared with after fibrinolysis. (Am Heart J 2006;151:798.e1-798.e7.) Fibrinolytic treatment of ST-elevation myocardial infarction (STEMI) is well established as reperfusion therapy for acute STEMI. The earlier the therapy is initiated, including the prehospital phase, the better the From the a Stockholm South Hospital, Stockholm, Sweden, b Sahlgrenska University Hospital/S, Gothenburg, Sweden c Hospital/Ö, Gothenburg, Sweden, and d The Brigham and Women s Hospital, Boston, Mass. This study was supported by grants from the Swedish Heart and Lung Foundation, Västra Götalands Foundation, Radi Medical AB Sweden, Laerdal Foundation and research grants from Roche Sweden AB, Aventis Sweden, Lilly International, and Infinity Medical Sweden. Submitted June 22, 2005; accepted September 14, Reprint requests: Leif Svensson, MD, PhD, Department of Cardiology, Karolinska Institute, Stockholm South Hospital, S Stockholm, Sweden. leif.svensson@sodersjukhuset.se /$ - see front matter n 2006, Published by Mosby, Inc. doi: /j.ahj outcome. 1-4 As an alternative, primary percutaneous coronary intervention (PCI) results in an immediate patency of the infarct-related artery in a high proportion of patients, 5-8 although the inherent time delay with this strategy may be related to adverse outcomes. 9,10 For both reperfusion strategies, even if antegrade flow is established, this restoration of flow in the epicardial infarct-related artery is not always accompanied by normal perfusion of the myocardium. 11 Adjunctive antiplatelet and antithrombotic therapy given before PCI may result in more complete reperfusion and thereby limit myocardial injury. 12 Several agents such as aspirin, heparin, fibrinolytic therapy, and more complete platelet inhibition with glycoprotein (GP) IIb/IIIa receptor antagonists, have been suggested as measures to improve outcomes when given before PCI as compared with primary PCI alone

2 798.e2 Svensson et al American Heart Journal April 2006 When studying reperfusion strategies in STEMI, mortality is considered a bhardq end point. Mortality studies, however, require the enrollment of vast numbers of patients. Vessel patency and flow in epicardial vessels (TIMI flow) have been associated with mortality and may therefore act as surrogate end points. 11 Recent observations indicate that resolution of the ST elevation, as a reflection of myocardial perfusion, is likewise significantly related to epicardial flow, myocardial perfusion, and mortality after reperfusion therapy. 16,17 Thus, resolution of ST elevation and TIMI flow may serve as meaningful effective, cheap, and easily assessable surrogate end points, thus reducing the needed population size in studies comparing different reperfusion strategies for acute myocardial infarction. The main objective of this trial was to compare a fibrinolytic strategy with an interventional strategy, both with up-to-date adjunctive pharmacological therapies initiated as early as possible preferably in the prehospital phase with respect to ST resolution 120 minutes after inclusion, and TIMI flow was evaluated at a mandated coronary angiography 5 to 7 days after inclusion. Methods Study organization and participants In total 7 hospitals (4 with and 3 without on-site cardiac laboratory facilities), and their affiliated ambulance organizations took part in the study. After a rapid clinical assessment and application of a checklist of inclusion and exclusion criteria, the patient was considered for enrollment. The decision to include and treat a patient in the study was always made by a physician after interpreting the actual electrocardiogram (ECG), either obtained prehospitally and sent by a cellular phone system to the randomizing hospital or taken on site in the hospital, and evaluating the reported checklist information. The local ethical committees and the Swedish Medical Products Agency approved the study. Study population Patients were included from November 2001 to May Inclusion criteria. Patients z18 years old presenting within 6 hours after onset of symptoms, with a duration of z30 minutes and with ST-segment elevation z2 mm in at least 2 adjacent precordial leads or z1 mm in at least 2 adjacent limb leads were considered for inclusion. Exclusion criteria. Initially, there was no upper age limit. However, after the results of the ASSENT 3 PLUS study 18 were disclosed, an upper limit of 75 years was introduced. Among other exclusion criteria were blood pressure more than 180/110 mm Hg at any time, known bleeding disorders, or any contraindication to fibrinolysis, cardiogenic shock defined as systolic blood pressure of b90 mm Hg together with cerebral deterioration and peripheral coldness, cardiopulmonary resuscitation lasting N10 minutes within 2 weeks before inclusion, ongoing treatment with anticoagulants such as low-molecular-weight heparin or warfarin (within 7 days unless international normalized ratio is V1.2), allergy to study drugs, body weight N120 kg, and known renal insufficiency (S-creatinine N200 mmol/l). Randomization and study protocol Randomization was stratified by centers and within the centers performed in blocks of 10 sealed envelopes to ensure balanced distribution of the 2 treatment groups at any time. This stratification was done before study start by an independent company. In the sealed envelope, treatment modality was documented as treatment A (fibrinolysis) or treatment B (invasive treatment). Randomization was done by opening a sealed envelope in consecutive order at the local including hospital. For patients in both groups, the study medications were administered as soon as possible, if applicable in the prehospital setting. If included prehospitally or at a hospital without on-site catheterization facility, patients in group A were taken to the nearest coronary care unit and treated according to current routines, and patients in group B were immediately transferred to a PCI center for coronary angiography and subsequential immediate PCI if appropriate. Study drug treatment Enoxaparin was administered as soon as possible after randomization to all patients. Patients in the fibrinolytic group received an intravenous bolus injection of 30 mg of enoxaparin followed by subcutaneous injection of the drug 1 mg/kg body weight with a maximum of 100 mg that was repeated every 12 hours during hospital stay or at the longest for 7 days. In group B, an intravenous bolus of enoxaparin 0.75 mg/kg body weight was administered, whereafter no further enoxaparin or other heparin was given. To patients in the fibrinolytic group, the enoxaparin injections were followed by reteplase administered as a double-bolus injection ( U) at 30-minute intervals. To patients in the invasive group, abciximab was started with an intravenous bolus of 0.25 mg/kg body weight as soon as possible after the enoxaparin injection. Subsequent to admission to the PCI laboratory, abciximab was continued with a 12-hour infusions of 10 lg/min. Aspirin was administered to all patients. Clopidogrel was started as soon as possible after PCI in group B patients who received a stent. All other medical therapy was given at the discretion of the physician in charge of the patient. Revascularization policies After the qualifying angiogram in the invasive group. An infarct-related artery was defined as an epicardial artery demonstrating a significant lesion localized in an artery supplying the ischemic area defined by ST elevations on the presenting ECG. If this lesion had a stenosis of z70% diameter and/or decreased antegrade flow (flow less than TIMI 3), and if the lesion was treatable with PCI, a PCI was performed, preferably with the use of stents. PCI was performed whether or not the patient at the time of angiography had ongoing ischemia. If the lesion was considered not appropriate for PCI, and if the patient had ongoing symptoms/ischemia, coronary artery bypass graft (CABG) surgery was to be considered.

3 American Heart Journal Volume 151, Number 4 Svensson et al 798.e3 If PCI was performed, and the coronary angiogram demonstrated a multivessel disease, it was left to the operators discretion to treat other than the culprit lesion if considered to be of clinical importance in an acute situation. Otherwise, nonculprit lesions were left and considered for a staged procedure in a stable phase. In the postinfarction period. As part of the protocol, coronary angiography was mandated in all patients before hospital discharge, preferably 5 to 7 days after randomization. Indications for revascularization in connection to this angiography or later were as follows: (1) Significant coronary artery disease to an extent that deemed revascularization necessary for prognostic reasons. This implies left main stenosis or 3-vessel disease including the stenosis in the proximal part of the left anterior descending coronary artery. (2) Significant stenosis/occlusion of the infarct-related artery in combination with either of the following: (a) episodes of angina in the postinfarction period, (b) signs of ischemia on ECG monitoring, or (c) signs of ischemia on an exercise test preceding the coronary angiography. (3) If the control angiography revealed significant coronary artery disease as described above, but the patient did not fulfill any of the other criteria before this angiography, a later decision for revascularization was taken if the patient was seen to have a disabling angina pectoris or signs of ischemia on a later exercise test or stress echocardiography. Thus, if the control angiography revealed an infarct-related artery with significant stenosis or occlusion, revascularization was not recommended in the absence of anginal symptoms or demonstrated ischemia. A coronary angiography was not performed after 5 to 7 days after infarction if the patient previous to that time had undergone a non-prespecified revascularization procedure as a result of failed fibrinolysis, early reinfarction, or ischemia after the initial treatment (fibrinolysis or PCI). Outside what was defined as part of the protocol. The use of rescue PCI after fibrinolytic treatment was left to the discretion of the treating physician. In general, at the participating hospitals, the following policy was applied: patients were considered for rescue PCI if, 90 minutes after the start of fibrinolysis, they had ongoing symptoms in combination with ST resolution of b30% of initial ST elevation in the worst lead at inclusion. Patients with initial symptom relief and ST resolution of z30% but with recurrent chest pain and ST elevations within 24 hours after randomization were also considered for this treatment. If the patient presented with postinfarction angina/ischemia, with or without recurrent ST elevations, representing an unstable situation that could not be controlled by medical treatment, the patient was referred for immediate angiography. In the postdischarge period, a new coronary angiography was performed if the patient had disabling angina or had demonstrable signs of ischemia on an exercise test or was readmitted with clinical signs of unstable angina or STEMI. Primary end points and efficacy evaluation Twelve-lead ECGs were to be recorded at randomization and 60, 90, 120, and 240 minutes after randomization. All ECGs were analyzed at the Ischemia Core Laboratory at Sahlgrenska University Hospital/Ö, Gothenburg, Sweden, with operators blinded to treatment allocation. Average ECG complexes or at least 2 QRS complexes were analyzed, and the mean ST elevation was noted. ST elevation was measured in the lead with the most prominent ST elevation on a 12-lead ECG at randomization, and the primary end point was defined as an ST resolution in this lead of at least 50% 120 minutes after randomization (allowing a F20 minutes discrepancy from the exact randomization and 120-minute time points). TIMI flow grade according to the TIMI flow classification 19 was evaluated on all coronary angiograms performed throughout the study. This analysis was performed by an angiographic core laboratory (TIMI/Perfuse Core Laboratory, Boston, MA) with operators blinded to treatment allocation. Primary analyses were done using the scheduled perprotocol angiographies, performed preferably 5 to 7 days after inclusion in the study. To compensate for patient dropouts as a result of 5- to 7-day angiographies not being routinely performed for patients undergoing rescue or subacute PCI during the first few days after inclusion, 2 additional analyses were conducted. (1) The first angiography performed after the initial randomized treatment was used; that is, for patients randomized to group A, the first angiography performed, regardless of reason, was used, and for group B, the first angiography performed after the initial primary procedure (excluding the postintervention angiography connected to this initial procedure), regardless of reason. (2) Control angiographies 5 to 7 days after inclusion in the study were used when available, and for those with no such angiography, the last performed angiography during the first 5 days after inclusion, regardless of reason, was used, excluding those angiographies performed in connection with initial randomized treatment in the invasive group; that is, in this analysis, the postprocedure angiographies, after rescue or subacute PCIs of patients with missing control angiographies, were included. Clinical and secondary end points All patients were followed up for 30 days after randomization. Death, myocardial infarction, and stroke, alone or as a composite during hospital stay and during the 30-day followup period, constituted secondary end points. Myocardial infarction was defined as 2 of the following: (1) chest pain indicative of myocardial infarction, (2) development of Q waves on ECG, and (3) elevation of markers of myocardial injury greater than normal. Stroke was defined as onset of neurological deficit and result of computed tomography scan or magnetic resonance imaging suggestive of either intracerebral hemorrhage or ischemic injury. The composite of death, myocardial infarction, stroke, and any revascularization after the index procedure until 30 days after inclusion constituted another secondary end point. Corrected TIMI frame count (CTFC) at the control angiographies served as a secondary end point. This evaluation was made in conjunction to evaluation of TIMI flow grade at the TIMI/Perfuse Core Laboratory. Statistical considerations Power and sample size considerations. The power calculations were made with respect to expected outcomes

4 798.e4 Svensson et al American Heart Journal April 2006 Table I. Baseline characteristics Invasive group (n = 101) Fibrinolytic group (n = 104) Age mean F SD (y) 65.3 F F 12.4 Men/women (%) 73/27 75/25 History of previous angina pectoris (%) Previous myocardial infarction (%) History of heart failure (%) 2 3 History of hypertension (%) Diabetes mellitus (%) Previous CABG (%) 4 2 Previous PCI (%) 6 7 Current smoker (5/5)* (%) Place of randomization Before hospital admission (%) In emergency department (%) 0 1 In coronary care unit (%) Anterior wall myocardial infarction (%) Table II. ST-resolution number of patients with ST resolution z50% at 4 different time points after randomization Invasive group Fibrinolytic group At 60 min (n = 66/72)* 31 (47) 26 (36).20 At 90 min (n = 67/69)* 35 (52) 38 (55).74 At 120 min (n = 75/74)* 51 (68) 47 (64).56 At 240 min (n = 61/66)* 56 (92) 57 (86).33 Results are given as n (%). *Number of analyzed patients in the 2 treatment groups, respectively. Figure 1 P *Number of patients where information was missing, in the 2 groups, respectively. regarding the 2 primary end points, ST resolution 120 minutes after initiation of the study, and TIMI 3 flow 5 to 7 days after start of the study. Concerning ST resolution, it could be anticipated from the TIMI 14 trial that approximately 60% of patients undergoing therapy with a combination of fibrinolysis and abciximab would have ST resolution N50% 120 minutes after initiation of therapy. 16 In the present study. we administered the combination of fibrinolysis and low-molecular-weight heparin, and we did not expect the ST resolution rate to be higher than the one previously mentioned, that is, a 40% failure with regard to this goal. Concerning direct PCI, data on ST resolution have been scarce, but based on the findings of Shah et al, 17 we anticipated to find at least 80% of ST resolution, that is, a 20% failure, if we could achieve N90% of TIMI 3 flow in this arm. 17 Thus, the power calculations were based on the aim to prove a 50% reduction of failure to achieve ST resolution, and with a power of 80% and a significant level of.05 (2-sided test), a total of 166 (2 83) patients would be required. Concerning TIMI flow data from the GUSTO I trial, RAPID II trial, and a meta-analysis by Barbegelata et al, we expected to find that approximately 60% of patients undergoing modern fibrinolytic therapy had TIMI 3 flow at an angiography 5 to 7 days after therapy. After interventional therapy with stenting and with adjunctive therapy with GPIIb-IIIa receptor antagonist, data from the ADMIRAL trial made us expect to find at least 90% of the patients with TIMI 3 flow. 15 Thus, a therapeutic failure of 30% in the thrombolytic arm and 10% in the interventional arm was expected. From this approximately 65% reduction of therapeutic failure and with a power of 90% and a significant level of.05 (2-sided test), a total of 180 (2 90) patients would be required. To allow for some losses at follow-up, we therefore, before the study, decided to randomize 200 patients. All power and sample size calculations were made before the start of the study. Numbers of patients with TIMI 3 flow at control angiographies (primary analysis) in group A (filled bars indicate fibrinolytic group) and in group B (hatched bars indicate invasive group). Data are given for the primary analysis and for the alternative analyses to compensate for the losses of control angiographies (see text). Statistics Continuous variables are presented as medians (with 25th and 75th percentiles) or as means F SDs (concerning age of the patients). All analyses were performed on an intention-totreat basis; that is, patients were analyzed according to the treatment group they were randomized to, regardless of actual treatment received. For dichotomous variables, the v 2 and Fisher exact tests were used to test for differences that were expressed as risk reduction (RR) with 95% CIs. For ordered/continuous variables, the Mann-Whitney U test was used. All P values are 2-sided and without correction for multiple comparisons. Results A total of 205 patients were included, with 104 assigned to the medical strategy (group A) and 101 to the primarily invasive group (group B).

5 American Heart Journal Volume 151, Number 4 Svensson et al 798.e5 Table III. Clinical events (secondary end points) Invasive group Fibrinolytic group Before discharge Within 30 d Before discharge Within 30 d P* Py Death (%) Reinfarction (%) Stroke (%) Any of the 3 above (%) Rescue PCI (%) b.0001 b.0001 Subacute PCI (%) b.0001 b.0001 CABG (%) Any of all 7 above (%) b.0001 b.0001 *P value for difference between groups before discharge. yp value for difference between groups within 30 days. Baseline characteristics and logistics There were none or marginal differences in the distribution of baseline characteristics between the 2 groups (Table I). Forty-one percent of the patients in the medical group and 43% in the invasive group received their treatment in the prehospital phase. The median delay between onset of symptoms and treatment (start of fibrinolysis or balloon dilatation) was 114 minutes (25th, 75th percentiles: 83, 197) in the fibrinolytic group and 202 minutes (154, 276) in the invasive group (P b.0001). Median time delay to start of abciximab in the invasive group was 126 minutes (90, 210). Among patients randomized before hospital admission, there were no deaths or bleedings, and only one cardiac arrest (ventricular fibrillation) occurred. ST resolution There were no statistically significant difference between the 2 groups in number of patients reaching the primary end point of N50% ST resolution at 120 minutes (PCI vs fibrinolysis: RR 1.07, 95% CI ; P =.56; Table II) (primary end point). Furthermore, at 240 minutes, there were also no significant differences between the 2 groups (PCI vs fibrinolysis: RR 1.06, 95% CI ; P =.33; Table II). TIMI flow and CTFC at control angiographies Actual median time to control angiographies was 5 days in the invasive group and 6 days in the fibrinolytic group. TIMI flow analyzed on per-protocol angiographies after 5 to 7 days showed significantly more patients with TIMI 3 flow in the invasive as compared with the fibrinolytic strategy group (PCI vs fibrinolysis: RR 1.31, 95% CI ; P =.04; Figure 1). Thus, 54% in the fibrinolytic group versus 71% in the invasive group had TIMI 3 flow. In the invasive group, TIMI 3 flow immediately after direct PCI was achieved in 76% of the patients. Evaluation of CTFC showed similar results with significantly lower values in the invasive than in fibrinolytic arm median 31 (25th, 75th percentiles 22, 41) vs 36 (27, 51); P =.03. Clinical events Death, reinfarction, and stroke rates were not significantly different between the 2 groups when analyzed separately (Table III). At 30 days, the composite of death, stroke, or reinfarction occurred in 8% in the fibrinolytic group compared with 3% in the invasive group (not significant). Rescue PCI was performed in 23% and subacute PCI in 27% of the fibrinolytic group. Thus, the composite of death, reinfarction, stroke, and any revascularization within 30 days occurred in 58% in group A versus 10% in group B ( P b.0001). Discussion Although the use of PCI as primary reperfusion strategy for STEMI is gaining wider acceptance based on a number of studies, 23 there is still some debate as to its value compared with very early and preferably prehospital-initiated fibrinolytic therapies. 3,24 There is also some uncertainty as to whether the use of modern potent adjunctive antithrombotic agents will affect the outcome of both thrombolytic and invasive therapies. 18,25-28 One of the main findings of this study was that no significant differences in terms of ST resolution, despite much more rapid start of reperfusion therapy in the fibrinolytic group could be demonstrated. One explanation may be the consequent use of abciximab as a facilitation before PCI. This therapy was initiated only 12 minutes later, after onset of symptoms, than fibinolysis in the noninvasive group. Data on adding abciximab to PCI with stent have in the recent ACE study raised the hope of substantially better reperfusion than what has previously been accomplished. 29 Thus, Antoniucci et al 29 achieved 85% ST resolution, N50% on ECGs taken 30 minutes after a completed PCI, if

6 798.e6 Svensson et al American Heart Journal April 2006 abciximab was given compared with 68% if no abciximab was given. Our figures were lower but an explanation can be that, in our study, the ECGs were taken at fixed times after inclusion. The second major finding was that TIMI 3 flow in the infarct-related artery was better preserved in the invasive group compared with the fibrinolytic group. Nevertheless, it is obvious that despite successful PCI performances and the use of stents and GPIIb-IIIa receptor antagonists, normal flow in the infarct-related epicardial vessels was seen in only 76% of the patients in this study immediately after completed procedures, and perhaps even more seldom after approximately 1 week. Microembolism, edema, and injured tissue, which prohibit adequate run off in the microvasculature, are some possible mechanisms. The poor outcome with regard to TIMI flow is in line with the small and nonsignificant difference regarding ST resolution in our study and may be one explanation why it has been so difficult to unequivocally prove superiority of PCI over fibrinolysis, despite the rather poor outcomes regarding vessel patency and myocardial perfusion after the latter therapy. Myocardial protection and, possibly, protection against reperfusion injury are fields that still need more care and intervention. The clinical outcomes in our study were generally good, with low mortality in both groups and very few reinfarctions within the first 30 days. There are several possible explanations for the low mortality rate. One is selection bias which is obvious because most patients in the study were hemodynamically stable upon inclusion. Another explanation may be the fact that N40% of the study population was recruited and already treated in the prehospital phase. A further explanation to the low mortality rates in this study may be the high numbers of interventions, rescue or subacute, before discharge in the fibrinolytic group. Thus, as in the CAPTIM study, 23% of the patients in our study had a rescue PCI after fibrinolysis, and another equally large proportion had an intervention because of ischemia or angina before discharge. 3 These results focus on the importance of access to catheterization laboratories and facilities for PCI around the clock even if fibrinolysis is chosen as the primary strategy. A further observation was that both the combination of a fibrin-specific fibrinolytic agent and the lowmolecular-weight heparin enoxaparin and, in the invasive group, mandatory abciximab in combination with enoxaparin worked without logistical problems and with adverse effects well within acceptable range. Limitations The amount of missing data regarding both ECG and angiographic evaluations is an obvious problem. For the angiographic analysis, where 5- to 7-day data were missing mainly as a result of interventions performed in the medical group on the first few days after randomization, we tried to compensate for this loss by using the angiographies from these additional interventions in the analysis. Because the result from this expanded analysis was well in line with a more conservative analysis of perprotocol scheduled control angiographies, we believe that the results could be considered as fairly robust. Concerning the ECG analysis, there were several reasons for missing data, including inappropriate timing, lost ECG recordings, bundle-branch block, and other complications making ECG analysis impossible. Conclusions In applying modern adjunctive medical treatments and putting effort into starting therapy as soon as possible after first contact with the health care system, a comparison between fibrinolysis and abciximab-facilitated primary PCI as a strategy for acute STEMI came out without any demonstrated difference with regard to ST resolution as a marker of myocardial perfusion. On the other hand, analysis of epicardial flow in infarct-related arteries at angiographies before hospital discharge showed a better outcome after PCI than after fibrinolysis. Still, when using an invasive strategy, the effects on flow in the infarct-related arteries were poorer than expected. Even if the rates of deaths and non procedure-related reinfarctions were low in both groups, the results indicate that more efforts in improvement of either therapy are warranted. Furthermore, if using a fibrinolytic strategy, it is obviously important to have access to catheterization laboratories with PCI facilities because, in this study, a large proportion of the patients had either a rescue PCI or a subacute PCI before discharge. We thank research nurses Katarina Bohm-Lindkvist, Tina Wolmeryd, Margaretha Schflin, and Monica Eriksson for their enthusiastic work and knowledge in GCP. References 1. Weaver WD, Cerqueira M, Hallstrom AP, et al. Prehospital-initiated vs hospital-initiated thrombolytic therapy: the myocardial infarction triage and intervention trial. JAMA 1993;270: Widimsky P, Budesinsky T, Vorac D, et al. Long distance transport for primary angioplasty vs immediate thrombolysis in acute myocardial infarction. Final results of the randomized national multicentre trial PRAGUE-2. Eur Heart J 2003;24: Bonnefoy E, Lapostolle F, Leizorovicz A, et al. Comparison of Angioplasty and Prehospital Thrombolysis in Acute Myocardial Infarction study group. Primary angioplasty versus prehospital fibrinolysis in acute myocardial infarction: a randomised study. Lancet 2002;360: Svensson L, Karlsson T, Nordlander R, et al. Implementation of prehospital thrombolysis in Sweden: components of delay until delivery of treatment and examination of treatment feasibility. Int J Cardiol 2003;88:

7 American Heart Journal Volume 151, Number 4 Svensson et al 798.e7 5. Grines C, Browne K, Marco J, et al. A comparison of immediate angioplasty with thrombolytic therapy for acute myocardial infarction. N Engl J Med 1993;328: Zijlstra F, de Boer M, Beukema W, et al. Mortality, reinfarction, left ventricular ejection fraction and costs following reperfusion therapies for acute myocardial infarction. Eur Heart J 1996;17: Widimsky P, Groch L, Zelizko M, et al. Multicentre randomised trial comparing transport to primary angioplasty vs immediate thrombolysis vs combined strategy for patients with acute myocardial infarction presenting to a community hospital without a catheterisation laboratory. Eur Heart J 2000;21: Schfmig A, Kastrati A, Dirschinger J, et al. Coronary stenting plus platelet glycoprotein IIb/IIIa blockade compared with tissue plasminogen activator in acute myocardial infarction. N Engl J Med 2000;343: De Luca G, Suryapranata H, Zijlstra F, et al. Symptom onset to balloon time and mortality in patients with acute myocardial infarction treated by primary angioplasty. J Am Coll Cardiol 2003;42: De Luca G, Suryapranata H, van t Hof AW, et al. Prognostic assessment of patients with acute myocardial infarction treated with primary angioplasty: implications for early discharge. Circulation 2004;109: van t Hof AW, Liem A, Suryapranata H, et al. Angiographic assessment of myocardial reperfusion in patients treated with primary angioplasty for acute myocardial infarction: myocardial blush grade. Circulation 1998;16: Montalescot G, Barragan P, Wittenberg O, et al. Platelet glycoprotein IIb/IIIa inhibitor with coronary stenting for acute myocardial infarction. N Engl J Med 2001;344: Ross AM, Coyne KS, Reiner JS, et al. A randomized trial comparing primary angioplasty with a strategy of short-acting thrombolysis and immediate planned rescue angioplasty in acute myocardial infraction: the PACT trial. J Am Coll Cardiol 1999;34: Zijlstra F, Ernst N, de Boer MJ, et al. Influence of prehospital administration of aspirin and heparin on initial patency of the infarct-related artery in patients with acute ST elevation myocardial infarction. J Am Coll Cardiol 2002;39: Montalescot G, Barragan P, Wittenberg O, et al. Abciximab before direct angioplasty and stenting in myocardial infarction regarding acute and long-term follow-up. Platelet glycoprotein IIb/IIIa inhibition with coronary stenting for acute myocardial infarction. N Engl J Med 2001;21: de Lemos J, Antman E, Gibson M, et al. Abciximab improves both epicardial flow and myocardial reperfusion in ST-elevation myocardial infarction. Circulation 2000;101: Shah A, Wagner G, Granger C, et al. Prognostic implications of TIMI flow grade in the infarct related artery compared with continuous 12-lead ST-segment resolution analysis. J Am Coll Cardiol 2000;35: Wallentin L, Goldstein P, Armstrong PW, et al. Efficacy and safety of tenecteplase in combination with the low-molecular-weight heparin enoxaparin or unfractionated heparin in the prehospital setting: the Assessment of the Safety and Efficacy of a New Thrombolytic Regimen (ASSENT)-3 PLUS randomized trial in acute myocardial infarction. Circulation 2003;108: Chesebro JH, Knatterud G, Roberts R, et al. Thrombolysis in Myocardial Infarction (TIMI) trial, phase I: a comparison between intravenous tissue plasminogen activator and intravenous streptokinase: clinical findings through hospital discharge. Circulation 1987;76: The GUSTO angiographic investigators. The effect of tissue plasminogen activator, streptokinase or both on coronary-artery patency, ventricular function and survival after acute stage of myocardial infarction. N Engl J Med 1993;329: Bode C, Smalling RW, Berg G, et al. Randomized comparison of coronary thrombolysis achieved with double-bolus reteplase (recombinant plasminogen activator) and front-loaded, accelerated alteplase (recombinant tissue plasminogen activator) in patients with acute myocardial infarction. Circulation 1996;94: Barbegelata NA, Granger C, Oqueli E, et al. TIMI grade 3 flow and reocclusion after intravenous thrombolytic therapy: a pooled analysis. Am Heart J 1997;133: Keeley EC, Boura JA, Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361: Steg PG, Bonnefoy E, Chabaud S, et al. Comparison of Angioplasty and Prehospital Thrombolysis In acute Myocardial infarction (CAPTIM) investigators. Impact of time to treatment on mortality after prehospital fibrinolysis or primary angioplasty: data from the CAPTIM randomized clinical trial. Circulation 2003;108: The Assessment of the Safety and Efficacy of a New Thrombolytic Regimen (ASSENT)-3 Investigators. Efficacy and safety of tenecteplase in combination with enoxaparin, abciximab, or unfractionated heparin: the ASSENT-3 randomized trial in acute myocardial infarction. Lancet 2002;358: Topol EJ, GUSTO V Investigators EJ. Reperfusion therapy for acute myocardial infarction with fibrinolytic therapy or combination reduced fibrinolytic therapy and platelet glycoprotein IIb/IIIa inhibition: the GUSTO V randomised trial. Lancet 2001;357: Kastrati A, Mehilli J, Dirschinger J, et al. Stent versus Thrombolysis for Occluded Coronary Arteries in Patients with Acute Myocardial Infarction (STOPAMI-2) Study. Myocardial salvage after coronary stenting plus abciximab versus fibrinolysis plus abciximab in patients with acute myocardial infarction: a randomised trial. Lancet 2002;359: Kereiakes DJ, Grines C, Fry E, et al. Enoxaparin and abciximab adjunctive pharmacotherapy during percutaneous coronary intervention. J Invasive Cardiol 2001;13: Antoniucci D, Rodriguez A, Hempel A, et al. A randomized trial comparing primary infarct artery stenting with or without abciximab in acute myocardial infarction. J Am Coll Cardiol 2003; 42: Appendix A Steering committee: Lars Grip, Leif Svensson, Mikael Aasa, Mikael Dellborg, 2ke Ohlsson, and Johan Herlitz. Safety monitor: Hans Öhlin, Lund University Hospital, Sweden. Statistician: Thomas Karlsson, Sahlgrenska University Hospital, Gothenburg, Sweden. Participants: Thomas Kellert, Örebro University Hospital, Henrik Fryklund, Norra Älvsborgs L7ns Hospital, Martin Risenfors, Sahlgrenska University Hospital/M, and Tommy Carlsson, Varberg Hospital.