IST
(1997)Objective
In acute ischaemic stroke, does early antithrombotic therapy — aspirin 300 mg daily and/or subcutaneous unfractionated heparin — reduce death within 14 days or death or dependency at 6 months?
Study Summary
• Death or dependency at 6 months with aspirin: 61·2% vs 63·5% (unadjusted 2p=0·06; adjusted for baseline severity, 14 [SD 6] fewer per 1000, 2p=0·03).
• Subcutaneous heparin (low + medium dose combined) did not change death at 14 days (9·0% vs 9·3%) or death or dependency at 6 months (62·9% vs 62·9%); fewer recurrent ischaemic strokes (2·9% vs 3·8%) were offset by more haemorrhagic strokes (1·2% vs 0·4%, 2p<0·00001).
• Medium-dose heparin (12 500 IU bd) caused more death or non-fatal stroke at 14 days than low-dose (10·8% vs 12·6%, 2p=0·007) and more transfused/fatal extracranial bleeds (2·0% vs 0·6%).
• Aspirin was associated with 5 (SD 1) extra transfused/fatal extracranial bleeds per 1000; heparin with 9 (SD 1) extra.
Intervention
Aspirin 300 mg daily and/or subcutaneous unfractionated heparin (5000 IU bd or 12 500 IU bd) for up to 14 days, started within 48 h of stroke onset
Inclusion Criteria
Acute stroke of any severity, onset <48 h; no evidence of intracranial haemorrhage; physician uncertain whether to give aspirin and/or heparin
Study Design
Arms: 2×3 factorial: aspirin 300 mg vs avoid aspirin × (heparin 5000 IU bd or 12 500 IU bd vs avoid heparin)
Patients per Arm: Aspirin 9720 vs Avoid aspirin 9715; Heparin 9717 (4861 low-dose, 4856 medium-dose) vs Avoid heparin 9718
Outcome
• Primary: death or dependency at 6 months — aspirin 61·2% vs 63·5% (unadjusted 2p=0·06; adjusted 2p=0·03); heparin 62·9% vs 62·9% (NS).
• Recurrent ischaemic stroke at 14 days — aspirin 2·8% vs 3·9% (2p<0·001); heparin 2·9% vs 3·8% (2p=0·005).
• Symptomatic haemorrhagic stroke at 14 days — aspirin 0·9% vs 0·8% (NS); heparin 1·2% vs 0·4% (2p<0·00001).
• Transfused/fatal extracranial bleeds — aspirin 1·1% vs 0·6% (2p=0·0004); heparin 1·3% vs 0·4% (2p<0·00001).
Clinical Question
In patients with suspected acute ischaemic stroke, does early treatment with aspirin 300 mg daily and/or subcutaneous unfractionated heparin (5000 or 12 500 IU twice daily) for up to 14 days reduce death within 14 days or death or dependency at 6 months?
Bottom Line
Early aspirin 300 mg daily started within 48 h of ischaemic stroke onset produced a small but real reduction in death or non-fatal recurrent stroke at 14 days and in death or dependency at 6 months (after adjustment), while subcutaneous heparin conferred no net clinical benefit at 6 months, with medium-dose heparin causing more early bleeding and adverse outcomes than low-dose.
Major Points
- Large factorial 2×3 trial: 19 435 patients from 467 hospitals in 36 countries randomised within 48 h of suspected acute ischaemic stroke.
- Aspirin 300 mg daily significantly reduced recurrent ischaemic stroke within 14 days (2·8% vs 3·9%, 2p<0·001) without a significant excess of haemorrhagic stroke.
- Death or dependency at 6 months with aspirin was 61·2% vs 63·5% (unadjusted 2p=0·06); after adjustment for baseline stroke severity the benefit was significant (14 [SD 6] per 1000, 2p=0·03).
- Subcutaneous heparin (low + medium combined) yielded no significant reduction in death or dependency at 6 months (62·9% vs 62·9%); fewer recurrent ischaemic strokes were offset by more haemorrhagic strokes.
- Medium-dose heparin (12 500 IU bd) was more hazardous than low-dose (5000 IU bd), with excess 14-day death or non-fatal stroke (12·6% vs 10·8%, 2p=0·007) and excess transfused/fatal extracranial bleeds (2·0% vs 0·6%).
- Combined with the concurrently reported Chinese Acute Stroke Trial (CAST, 21 106 patients), aspirin prevented ~10 deaths or recurrent strokes per 1000 during the first few weeks after acute ischaemic stroke.
- IST supported starting aspirin as soon as ischaemic stroke was suspected once intracranial haemorrhage was excluded, and recommended against subcutaneous heparin regimens more intensive than 5000 IU bd.
Study Design
- Study Type
- Randomised controlled trial (open, factorial)
- Randomization
- Yes
- Blinding
- Open-label treatment; central 6-month outcome assessment blinded to allocation where done
- Sample Size
- 19435
- Follow-up
- 14 days in hospital and 6 months
- Centers
- 467
- Countries
- UK, Italy, Switzerland, Poland, Netherlands, Sweden, Norway, Argentina, Australia, Portugal, Spain, New Zealand, Czech Republic, Turkey, Belgium, Austria, India, Greece, Singapore, Hong Kong, USA, Canada, Israel, Hungary, Brazil, Slovak Republic, Slovenia, South Africa, Chile, Denmark, Ireland, Finland, Japan, Romania, Sri Lanka, France
Primary Outcome
Definition: Death from any cause within 14 days and death or dependency (needing help from another person with daily activities) at 6 months
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| - | - | - | aspirin 6-mo adjusted 2p=0·03; heparin 6-mo NS |
Limitations & Criticisms
- Open-label treatment without placebo control; potential for ascertainment bias in unblinded in-hospital events, though central 6-month follow-up was largely blinded.
- Pulmonary embolism appeared under-reported (0·7% overall vs 3-39% in published reviews), limiting inferences about heparin's VTE benefit.
- Broad eligibility (uncertainty principle) with limited pre-randomisation CT in 33%, and rapid administrative diagnosis instead of a formal severity score (e.g., NIHSS).
- Fixed subcutaneous heparin dosing without routine APTT monitoring; no intravenous heparin arm.
- Wide 48-h enrolment window with a median delay of 19 h means most patients were treated well beyond the early hyperacute period, limiting inference for very-early treatment.
- 6-month outcome collected largely by mailed questionnaire or phone in most countries (using the validated Warlow simple questions).
Citation
Lancet. 1997 May 31;349(9065):1569-1581. DOI: 10.1016/S0140-6736(97)04011-7. PMID: 9174558