We review GISSI-2 (Gruppo Italiano per lo Studio della Sopravvivenza nell'Infarto Miocardico [2, 3]), ISIS-3 (Third International Study of Infarct Survival [4]), and GUSTO-1 [1] to provide a rational basis for health providers to determine optimal patient care and resource allocation.

Currently, three thrombolytic regimens are approved by the Food and Drug Administration (FDA) for use in patients with acute myocardial infarction; two of these regimens were directly compared in GISSI-2, and all were directly compared in ISIS-3.

In GISSI-2 and its International Study Group extension, 20 891 patients with acute myocardial infarction who were admitted to coronary care units within 6 hours of symptom onset were enrolled in a randomized trial comparing the benefits and risks of streptokinase (1.5 million U over 30 to 60 minutes) and tissue plasminogen activator (tPA) (alteplase, 100 mg over 3 hours). All patients were given aspirin (325 mg) daily for 30 days and were also randomly assigned to receive either heparin (12 500 IU subcutaneously twice daily starting within 12 hours of thrombolysis) or no heparin.

Overall, no significant differences were observed in 35-day mortality rates between tPA and streptokinase (9.6% compared with 9.2%) or between heparin and no heparin (9.3% compared with 9.4%). In addition, no significant differences were observed among groups regarding major cardiac complications. A small and statistically significant excess of total strokes was, however, associated with tPA (1.33% compared with 0.94%, P = 0.008), as was an apparent but nonsignificant excess of cerebral hemorrhage (0.42% compared with 0.29%, P = 0.10). Long-term follow-up of patients in GISSI-2 showed no significant changes in the differences between tPA and streptokinase for any of these main clinical findings.

The Third International Study of Infarct Survival was the only randomized trial to directly compare all three thrombolytic agents: streptokinase, standard-dose tPA, and anisoylated plasminogen activator complex (APSAC). In brief, 41 299 patients who were admitted within 24 hours of the onset of the signs and symptoms of suspected acute myocardial infarction and with no known contraindication to thrombolytic therapy were randomly assigned to streptokinase (1.5 million U over 1 hour), tPA (duteplase, 0.60 million U/kg body weight over 4 hours), and APSAC (30 U over 3 minutes). All patients received aspirin (162.5 mg) daily. In addition, patients were randomly assigned to receive heparin (12 500 IU subcutaneously twice daily for 7 days starting 4 hours after randomization) or not to receive heparin.

The ISIS-3 trial showed no differences in mortality rates among thrombolytic agents. Rates for major in-hospital clinical events, including cardiogenic shock, heart failure requiring treatment, ventricular fibrillation, and cardiac rupture, were also virtually identical. A small but statistically significant deficit of in-hospital re-infarctions was observed in the tPA group. Not surprisingly, the bacterially derived proteins streptokinase and APSAC were associated with higher rates of allergy and hypotension requiring treatment, although the absolute rates for all three agents were exceptionally low in this double-blinded trial. Patients assigned to streptokinase had significantly fewer noncerebral bleeds than did those receiving either tPA or APSAC.

All three agents tested in ISIS-3 were associated with generally acceptable rates for cerebral complications; however, as in GISSI-2, patients assigned to streptokinase had fewer total strokes and intracranial bleeds. Although the absolute rates of cerebral hemorrhage in ISIS-3 were low for all three agents, the differences were statistically significant (P < 0.0001 for streptokinase compared with APSAC; P < 0.00001 for streptokinase compared with tPA).

Altogether in ISIS-3 and GISSI-2, more than 47 000 patients were randomly assigned directly to receive streptokinase or standard-dose tPA. Although ISIS-3 used a double-chain form of tPA (duteplase) and GISSI-2 used a predominantly single-chain form (alteplase), the results of each trial are identical with respect to mortality and major side effects. In fact, when the ISIS-3 and GISSI-2 data on streptokinase and tPA are combined, the mortality rate associated with each agent at 35 days is 10%. However, rates for total stroke (1.4% compared with 1.0%, P < 0.001) and for cerebral hemorrhage (0.6% compared with 0.3%, P < 0.00001) were significantly higher for tPA than for streptokinase.

Neither GISSI-2 nor ISIS-3 used intravenous heparin as adjunctive therapy to thrombolysis, an issue that has raised concerns about the generalizability of these trials to routine care in the United States. However, in data-derived subgroup analyses of patients from ISIS-3 who received intravenous heparin as a protocol deviation, no differential mortality effect was found when one agent was compared with another, although the stroke rate among those receiving intravenous heparin was nearly twice that among patients treated with subcutaneous heparin, regardless of whether they received streptokinase, tPA, or APSAC [5]. Such data are useful to raise questions but not to test hypotheses.

The GUSTO-1 trial evaluated two new "accelerated" tPA regimens for acute myocardial infarction and compared them with streptokinase given with either intravenous or subcutaneous heparin. Randomization to intravenous or subcutaneous heparin was not done for either of the new and currently unapproved tPA regimens, and APSAC was not tested.

In brief, GUSTO-1 randomly assigned 41 021 patients within 6 hours of onset of acute myocardial infarction to receive one of the following four treatments: accelerated-dose tPA combined with intravenous heparin, accelerated-dose tPA combined with streptokinase and intravenous heparin, standard-dose streptokinase combined with intravenous heparin, and standard-dose streptokinase combined with subcutaneous heparin (the regimen used in ISIS-3 and GISSI-2). All patients also received aspirin in a dose adequate to achieve a rapid clinical antithrombotic effect.

The primary clinical findings of the GUSTO trial were that 1) standard-dose streptokinase plus intravenous heparin had no advantage in terms of combined mortality and nonfatal stroke rates when compared with standard-dose streptokinase plus subcutaneous heparin [8.2% compared with 7.9%]; 2) accelerated-dose tPA combined with streptokinase and intravenous heparin had no beneficial effect for these end points when compared with standard-dose streptokinase plus subcutaneous heparin [7.9% compared with 7.9%]; and 3) accelerated-dose tPA plus intravenous heparin had a 0.9% mortality advantage over standard-dose streptokinase plus subcutaneous heparin which, because of the higher stroke rates associated with tPA (1.55% compared with 1.22%), resulted in an absolute reduction of 0.7% in the combined rate of death plus nonfatal stroke (7.9% compared with 7.2%). Although small in absolute magnitude, this difference was statistically significant. However, this finding was restricted to patients treated within 4 hours of symptom onset. Although the median time to treatment for patients enrolled in the GUSTO-1 trial was 2.8 hours, most patients with acute myocardial infarction in the United States come to the hospital more than 4 hours after onset of symptoms [6]; among patients treated in the GUSTO-1 trial who were admitted more than 4 hours after symptom onset, no thrombolytic regimen showed an advantage in terms of mortality, a finding consistent with the ISIS-3 and GISSI-2 results. In fact, for the more than 9000 patients in GUSTO-1 who were treated 4 or more hours after symptom onset, mortality was nonsignificantly lower among patients receiving streptokinase with subcutaneous heparin compared with those receiving accelerated-dose tPA with intravenous heparin.

The use of an open-label design in GUSTO-1 increases the difficulty of interpreting the true benefit-to-risk ratios associated with each of the newer and as-yet-unapproved thrombolytic regimens. For example, patients treated with accelerated-dose tPA were significantly more likely to have potentially life-saving emergent coronary artery bypass surgery [7] than were patients treated with streptokinase (9.5% compared with 8.5%), even though rates of congestive heart failure, pulmonary edema, and cardiogenic shock appeared to be somewhat lower in patients receiving accelerated-dose tPA than among those receiving streptokinase. Further, the benefit of accelerated-dose tPA observed in the GUSTO-1 trial was restricted to U.S. patients. Among the 17 803 patients in GUSTO-1 who were studied outside the United States, those receiving the combined regimen of accelerated-dose tPA plus streptokinase had the lowest mortality rate, whereas virtually no difference was found between the net clinical benefit associated with accelerated-dose tPA plus intravenous heparin when compared with standard-dose streptokinase plus subcutaneous heparin. Thus, the possibility of bias inadvertently introduced by the open-label design of GUSTO-1 cannot be excluded.

Several clinically relevant conclusions can be drawn from the randomization of more than 100 000 patients in these three large-scale trials directly comparing thrombolytic agents. First, findings from GISSI-2, ISIS-3, and GUSTO-1 consistently indicate that the choice of thrombolytic therapy is much less important to ultimate survival than is the delay time between onset of symptoms and initiation of treatment. Because the average in-hospital delay for patients treated with thrombolytic agents is almost 90 minutes in the United States [8], the development of local programs in emergency departments that are designed to decrease the delay time to initiation of thrombolytic therapy is probably the most cost-effective way to save the greatest number of lives. Second, potential differences in efficacy and safety among the available thrombolytic agents are at most small in absolute magnitude and are unlikely to pertain to most patients with myocardial infarction who come to the hospital more than 4 hours after the onset of pain. Because all three agents appear to be effective even when given up to 12 hours after the onset of symptoms, a clinical strategy must be adopted to increase thrombolytic use for late arrivers, regardless of which agent is chosen. Patients excluded from thrombolytic therapy because of age, sex, presence of bundle-branch block on the admission electrocardiogram, and late arrival in the emergency department do, in fact, benefit substantially from thrombolysis as long as contraindications to the drugs do not exist [9]. However, fewer than one third of U.S. patients with acute myocardial infarction receive thrombolytic therapy [10]. Thus, from the perspective of practicing physicians, any potential small differences among agents in efficacy, safety, and ease of administration must be recognized as being of far less clinical importance than is the wider use of thrombolytic therapy with any of the available agents.