Current issues of ACP Journal Club are published in Annals of Internal Medicine


Streptokinase led to higher 10-day and 6-month case fatality rates in acute stroke

ACP J Club. 1996 May-June;124:60. doi:10.7326/ACPJC-1996-124-3-060

Source Citation

Multicentre Acute Stroke Trial-Italy (MAST-I) Group. Randomised controlled trial of streptokinase, aspirin, and combination of both in treatment of acute ischaemic stroke. Lancet. 1995 Dec 9;346:1509-14.



To determine the effectiveness of streptokinase, aspirin, and a combination of both in treating patients with acute ischemic stroke.


Randomized controlled trial with 6-month follow-up.


70 centers in Italy, Portugal, and the United Kingdom.


622 patients (78% aged > 60 y, 54% men) who were admitted to the hospital after the sudden development of a focal neurologic deficit attributable to stroke, within 6 hours of symptom onset, and with no clear indication for or contraindication to streptokinase or aspirin. Exclusion criteria were intracerebral hemorrhage, severe coma without any purposeful motor response, rapid resolution of neurologic symptoms, or any medical disorders or exposure to drugs that might interfere with the safety or efficacy evaluation. Follow-up was complete.


Patients were randomized with a 2 × 2 factorial design: 157 to a 1-hour intravenous infusion of 1.5 MU of streptokinase alone; 156 to the same dose of streptokinase plus 300 mg/d of buffered aspirin for 10 days; 153 to 300 mg/d of aspirin alone; and 156 to neither drug.

Main Outcome Measures

Case fatality and disability.

Main Results

Streptokinase (alone or with aspirin) led to higher 10-day and 6-month case fatality rates than did aspirin alone or neither drug (27% vs 12% and 36% vs 24%, respectively, P < 0.001) but led to a lower 6-month disability rate (27% vs 40%, P < 0.001). Aspirin (alone or with streptokinase) did not increase the risk for early death when compared with streptokinase alone or neither drug (22% vs 16%, P > 0.05). Patients who received both streptokinase and aspirin had a higher 10-day case fatality rate than those who did not receive either drug (34% vs 13%, P < 0.001). Streptokinase alone did not increase the risk for early death when compared with neither drug (19% vs 13%, P = 0.12). Streptokinase plus aspirin led to a higher 6-month case fatality rate than did neither drug (44% vs 29%, P < 0.01) but to a lower 6-month disability rate (20% vs 39%, P < 0.001). No differences existed between streptokinase alone, streptokinase plus aspirin, and aspirin alone compared with neither drug in the incidence rates of combined 6-month case fatality and disability.


In patients with acute stroke, streptokinase with or without aspirin led to higher 10-day and 6-month case fatality rates than did neither drug or aspirin alone.

Sources of funding: In part, Pierrel SpA, Italy and Pharmacia Therapeutics, Sweden.

For article reprint: Professor L. Candelise, Istituto di Clinica Neurologica, Via F Sforza 35, 20122 Milano, Italy. FAX 39-2-5519-0392.


The MAST-I study arguably has generated the least discussion of the 3 large trials that published formal reports about intravenous thrombolytic therapy in acute stroke. Based on the positive findings of the National Institute of Neurological Disorders and Stroke (NINDS) rt-PA Stroke Study (1) and the neutral but encouraging results of the European Cooperative Acute Stroke Study (ECASS) (2), the notion has gained momentum that the time for the use of intravenous thrombolysis is at hand. Although published almost simultaneously with the NINDS rt-PA Study, MAST-I imparts a different message. This study showed that intravenous thrombolysis was too dangerous in the short term (within 10 days of starting treatment) to provide a statistically significant net benefit in the long term. The logical question that arises is whether intravenous thrombolysis for stroke will resemble the results from the NINDS rt-PA Study or MAST-I in the "real world."

The design features of MAST-I may be the source of the high early intraparenchymal bleed and mortality rates (e.g., a longer window for treatment and assignment of half the treated patients to aspirin). Streptokinase at 1.5 × 106 units may be more dangerous than rt-PA at 0.9 mg/kg. These features can be fixed for future studies. A more serious concern is how MAST-I was conducted by the scores of investigators at the 70 centers. By contrast, most patients who enrolled in the NINDS rt-PA Study were randomized by an elite cadre of experienced lytic therapists. Which model will most resemble the real world? In the ECASS, which was done more like MAST-I, the primary outcome analysis was prespecified, and it excluded patients who were misrandomized. Hence, 17% of the patients in the ECASS were excluded from the explanatory analysis because they had intrinsically high risk for bleeding or death. MAST-I included no such provision, nor would the real world.

MAST-I provides an important counterpoint to NINDS rt-PA Study by emphasizing the narrow window for thrombolysis and mandating that the safety of thrombolysis in real-world clinical practice be carefully monitored.

David C. Anderson, MD
Hennepin Country Medical CenterMinneapolis, Minnesota, USA


1. The National Institute of Neurologic Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med. 1995;333:1581-7.

2. Hacke W, Kaste M, Fieschi C, et al. JAMA. 1995;274:1017-25.