Review of the ISIS-2 Trial

First Mega-Trial of Aspirin and Revascularization in Patients with Suspected Acute Myocardial Infarction

Lancet 1988;349-360

Background The introduction to the ISIS-2 manuscript opens with a remarkable statement from the authors:

Reductions in mortality that are realistically moderate (eg, “only” 20-25%) are important, especially if produced by widely practicable treatments for common causes of death.

At the time of this posting, January 2024, it’s amazing to think, how in 1988, a 20-25% reduction in mortality was considered moderate for a “widely practicable treatment for a common cause of death” and now the profession seems to grasp at almost anything, at nearly any cost, that reduces some composite of hard, soft and, in some cases, completely meaningless endpoints by the same amount.

Regardless, ISIS-2 sought to test the hypothesis that streptokinase and aspirin, either alone or together, would reduce vascular mortality in patients with MI.

Patients Patients with suspected myocardial infarction, who the responsible physician thought were within 24 hours of symptom onset and had no clear indication for, or contraindication to, streptokinase or aspirin. Absolute contraindications included: any history of stroke or of gastrointestinal ulcer. Possible contraindications included: recent arterial puncture, recent severe trauma, severe persistent hypertension, allergy to streptokinase or aspirin, low risk of cardiac death, or some other life-threatening disease. ECG changes at entry were not required.

The design was meant to be pragmatic to facilitate patient enrollment; however, exclusion criteria was more extensive for ISIS-2 compared to ISIS-1, which reflects concern about hemorrhagic side effects associated with use of these agents. Like ISIS-1, there were no special procedures dictating patient follow-up after the hospitalization ended.

Baseline characteristics Basic demographic information is not provided in the main manuscript but we can infer from the subgroup Forest plots that more than two thirds of patients were men, the overwhelming majority were less than 70 years of age, and nearly half were less than 60. Less than 20% had a previous MI and only around 7% had diabetes. Patients with inferior and anterior STEMI’s composed more than half of the cohort. About 43% of patients presented within 4 hours of symptom onset, another 42% presented between 5 and 12 hours, and the remaining 15% presented between 13 and 24 hours.

Procedures A 2x2 factorial study design was used. All patients were randomly assigned to receive either streptokinase or a matching placebo. All patients were also randomly assigned to receive either aspirin or a matching placebo. This led to 4 distinct treatment groups: 1) streptokinase + aspirin placebo, 2) aspirin + streptokinase placebo, 3) streptokinase + aspirin, or 4) placebo only

Patients allocated to receive streptokinase were immediately given 1.5 MU of ‘Streptase’ over 1 hour. Patients allocated to receive aspirin were immediately given 162.5 mg of an enteric coated tablet that was crushed, sucked, or chewed for a rapid anti-platelet effect. Thereafter, they took a 162.5 mg tablet daily for 1 month.

Endpoints Vascular mortality over 5 weeks was the primary endpoint. Follow-up after discharge involved only mortality, through government records wherever possible.

Results 17,187 participants were randomized from 417 hospitals in 16 countries. Compliance with the assigned treatments was estimated to be between 90-95%.

The results of ISIS-2 are not reported in a standard results table, making vascular mortality as well as all-cause mortality impossible to infer over the duration of the trial. The number of non-vascular deaths are presented in a table but vascular deaths beyond 5 weeks are not formally provided in the original publication. The only evidence of vascular deaths beyond 5 weeks is presented in survival curves.

Over 5 weeks, Streptokinase alone reduced vascular mortality compared to placebo by approximately 25% (9.2% vs 12.0%) and these results were highly significant. The difference remained highly significant over the course of the trial. Non-vascular deaths accounted for a tiny fraction of all deaths (3.5%) and were unchanged between groups.

Subgroup analysis of streptokinase efficacy based on time from pain onset showed it was more effective if given within 4 hours (8.2% vs 12.3%) vs between 5-24 hours (10.0% vs 11.8%).

Streptokinase use was associated with higher rates of hypotension and bradycardia (10% vs 2.0%), allergic reactions (4.4% vs 0.9%), minor bleeding (3.5% vs 1.0%) and major bleeding (0.5% vs 0.2%). The bleeding excess appeared similar whether streptokinase was used with aspirin or not.

Streptokinase was also associated with a small but significant risk of cerebral hemorrhage (7 vs 0 events) all of which were followed by death or severe disability. However, overall, streptokinase was not associated with an increase in stroke (0.7% vs 0.8%) or in the risk of disabling/fatal stroke (0.5% vs 0.6%).

Aspirin alone reduced vascular mortality compared to placebo by 23%, approximately the same amount as streptokinase alone over 5 weeks (9.4% vs 11.8%), and these results were highly significant and remained so over the course of the trial. There were fewer non-vascular deaths among aspirin allocated patients, so all-cause deaths were also significantly reduced but like with streptokinase, non-vascular death accounted for a tiny fraction of all deaths. Unlike Streptokinase alone, the treatment effect of aspirin was unchanged regardless of when it was started relative to the onset of pain.

Aspirin significantly increased minor bleeds (2.5% vs 1.9%) but there were no major differences in any other side effects.

Streptokinase and aspirin compared to double placebo reduced vascular mortality nearly twice as much as either agent alone (8.0% vs 13.2%) and these results were highly significant and remained so over the course of the trial. Non-vascular deaths were the same between groups and so all-cause mortality was also significantly reduced by streptokinase and aspirin. Subgroup analysis based on time from pain onset demonstrated greater efficacy for those receiving treatment in under 4 hours (6.4% vs 13.1%) compared to those receiving treatment between 5-24 hours (9.2% vs 13.3%).

Streptokinase and aspirin together were associated with increases in major and minor bleeding as well as cerebral hemorrhage that were similar to streptokinase alone both in terms of their relative and absolute differences.

Additional subgroup analyses suggest that patients presenting with anterior STEMI’s derived the greatest benefit from streptokinase alone and in combination with aspirin. The risk of death was about twice as high for patients presenting with an anterior vs inferior STEMI (15% vs 8%). Patient’s presenting with ST depression on their ECGs experienced a similar overall risk of death compared to patients with anterior STEMI’s but did not appear to derive a significant benefit from any combination of treatment.

Conclusions Streptokinase alone and aspirin alone reduced death over 5 weeks compared to placebo. 30 to 40 patients would need to be treated with either agent alone to prevent 1 person from dying, which was very similar to the effect noted in the GISSI trial that tested Streptokinase alone.

In ISIS-2, the combination of the 2 agents (Streptokinase and Aspirin) reduced death even further (nearly twice the effect of either agent alone) compared to double placebo controls and this was associated with an NNT of 20 or less.

Streptokinase, whether alone or in combination with aspirin, exerted the greatest effect when given to patients presenting within 4 hours from pain onset but was still effective when given outside this window (subgroup stratification was not the same as GISSI, which showed that the benefit waned and possibly reversed after 6-9 hours).

In ISIS-2 there was no evidence to suggest that sicker patients or those with larger MI’s did worse (i.e, lower SBP, higher HR, or anterior STEMI) but again, not all subgroups presented were the same as in the GISSI trial, which did show less benefit for some higher risk groups. Like GISSI, ISIS-2 provided evidence that patients presenting with NSTEMI’s did not benefit from streptokinase.

In summation, ISIS-2, like the GISSI trial that preceded it, demonstrated the significant impact of early revascularization via thrombolysis on mortality in patients presenting with STEMI. It also demonstrated the significant benefit of aspirin and the additive effect of both agents together.

To this day, the pillars of STEMI management involve aspirin and prompt revascularization, whether done via thrombolysis or PCI.