CAST: Chinese Acute Stroke Trial: Aspirin 160 mg/day vs Placebo Within 48h of Acute Ischaemic Stroke (CAST Collaborative Group, 1997)
In patients within 48 hours of suspected acute ischaemic stroke, does aspirin 160 mg/day vs matched placebo reduce in-hospital mortality and death or dependence at discharge? Double-blind placebo-controlled trial in 21,106 Chinese patients — the rigorous-blinding counterpart to the open-label IST.
CAST Collaborative Group (Lancet 1997;349:1641-1649) · doi:10.1016/S0140-6736(97)04010-5 · 21,106 patients
Population
Included
- Suspected acute ischaemic stroke with onset within 48 hours before randomisation
- No clear contraindication to aspirin (recent gastric bleeding, known aspirin allergy)
- Responsible physician uncertain whether or not aspirin treatment was indicated for this particular patient (uncertainty principle)
- CT before randomisation mandatory only for comatose patients; other patients could be entered when the responsible physician was reasonably certain haemorrhagic stroke could be excluded
- Patients admitted to the participating hospitals in China were eligible if they were judged to be within 48 hours of the onset of symptoms of suspected acute ischaemic stroke, and had no clear indications for, or contraindications to, aspirin
- The fundamental criterion for entry was that the responsible physician was uncertain whether or not aspirin treatment was indicated for a particular patient
- A CT scan before randomisation was mandatory only for patients who were comatose; other patients could be entered for randomisation without a CT scan provided that the responsible physician was reasonably certain that haemorrhagic stroke could be excluded
Excluded
- Recent gastric bleeding or known aspirin allergy
- Only a minor stroke with little likelihood of worthwhile in-hospital benefit
- Other major life-threatening disease
- Severe pre-existing disability
- Contraindications based on an expectation of an increased risk of adverse effects, for example because there was a recent history of serious gastric bleeding or a known allergy to aspirin
- Little likelihood of any worthwhile benefit in hospital, for example because the patient had only a minor stroke, other major life-threatening disease, or severe pre-existing disability
Source: CAST (Chinese Acute Stroke Trial) Collaborative Group, Lancet 1997;349:1641-1649· Retrieved 2026-06-09
Primary Outcome
4-week in-hospital mortality (co-primary)
Negligible absolute difference
4-week in-hospital mortality
Combined 4-week death or non-fatal stroke: 5.3% vs 5.9% (2p=0.03; 6.8 fewer per 1000). Dead-or-dependent at discharge directionally favoured aspirin (30.5% vs 31.6%, 2p=0.08; trend). Small non-significant excess of haemorrhagic stroke (+2/1000) and significant excess of major extracranial bleeding (+2.7/1000, 2p=0.02). Pooled IST+CAST (Chen 2000): ~9/1000 fewer dead-or-restroked vs ~2/1000 excess ICH; net ~7/1000 benefit.
Study Arms
- Agent
- Aspirin (acetylsalicylic acid), film-coated tablet
- Dose
- 160 mg daily
- Route
- Oral (crushed or chewed for the first dose; nasogastric tube if the patient could not swallow tablets)
- Frequency
- Once daily
- Duration
- Up to 4 weeks, or until earlier discharge or death
- Co-interventions
- Non-trial aspirin and other antiplatelet drugs were not permitted during the trial period unless a strong indication developed (for example, acute myocardial infarction). Chinese herbal products were permitted at clinician discretion. The protocol encouraged long-term low-dose aspirin at discharge
Aspirin arm from the CAST Collaborative Group (Lancet 1997 Methods). Started immediately after randomisation. Aspirin produced a small but definite net benefit: in-hospital mortality 3.3% vs 3.9% (2p=0.04) and combined 4-week death or non-fatal stroke 5.3% vs 5.9% (2p=0.03). Fewer recurrent ischaemic strokes (1.6% vs 2.1%) with a slight nonsignificant excess of haemorrhagic strokes (1.1% vs 0.9%). Taken with IST, supports early aspirin in acute ischaemic stroke.
- Agent
- Matching placebo, identical tablets and packaging
- Route
- Oral (identical administration to aspirin arm)
- Frequency
- Once daily
- Duration
- Up to 4 weeks, or until earlier discharge or death
- Co-interventions
- Identical restriction on non-trial antiplatelet drugs and identical permission for Chinese herbal products as the aspirin arm
Double-blind, placebo-controlled. Randomisation used prepacked, sequentially numbered calendar packs of either aspirin or matching placebo, balanced for every ten consecutive patients within each hospital. Prospectively planned for parallel analysis with the International Stroke Trial (IST). Source: CAST Collaborative Group, Lancet 1997.
Safety
4-week in-hospital mortality (primary)
3.3%
3.9%
Aspirin 3.3% (343/10,554) vs placebo 3.9% (398/10,552), 2p=0.04. 14% (SD 7) proportional reduction; 5.4 (SD 2.6) fewer deaths per 1000.
4-week transfused or fatal extracranial bleeds
0.8%
0.6%
Aspirin 0.8% vs placebo 0.6% (2p=0.02; +2.7 per 1000). Restricted-to-survivors analysis: +1.9 per 1000 transfused bleeds (47 vs 27 events; 0.5% vs 0.3%, 2p=0.02).
4-week haemorrhagic stroke
1.1%
0.9%
Aspirin 1.1% vs placebo 0.9% (2p>0.1). Small non-significant excess of haemorrhagic stroke (+2 per 1000). The background risk of haemorrhagic stroke in the CAST Chinese population (9 per 1000 placebo arm) was higher than in IST (3 per 1000 avoid-aspirin/avoid-heparin reference), but the absolute excess associated with aspirin was similar in both trials.
Trial Design
Type
- Large, randomised, double-blind, placebo-controlled trial
- Central prepacked calendar-packed envelopes with sequential numbering; allocation balanced for every ten consecutive patients within each hospital
- 413 Chinese hospitals; recruitment began with a 600-patient pilot phase in 10 hospitals
- Intention-to-treat analysis; pre-specified to be analysed in parallel with the concurrent International Stroke Trial (IST)
- Interim analyses planned after 5,000, 10,000, and 15,000 patients; September 1996 interim showed promising results from both CAST and IST but the steering committee continued enrolment to scheduled end
- Two-sided p values (2p) cited throughout; absolute differences expressed as benefits per 1000 patients treated
Timeline
Enrolled November 1993 to March 1997; in-hospital outcome (up to 4 weeks); discharge forms available for 97.9% of randomised patients. Published Lancet 1997 June 7.
N
21,106
Enrollment
21,106 patients at 413 Chinese hospitals. Enrolled November 1993 to March 1997. Double-blind placebo-controlled; central calendar-packed envelopes; allocation balanced per 10 consecutive patients per hospital. In-hospital outcome up to 4 weeks; discharge forms for 97.9% of randomised patients. Pre-specified for parallel analysis with IST. Published Lancet 1997 June 7.
Bedside Pearl
CAST (with IST): double-blind placebo-controlled trial of aspirin 160 mg/day vs placebo in 21,106 patients within 48h of acute ischaemic stroke. In-hospital mortality 3.3% vs 3.9% (2p=0.04); combined death or non-fatal stroke 5.3% vs 5.9% (2p=0.03). About 2 per 1000 excess major extracranial bleeds; small non-significant excess of haemorrhagic stroke. Together with IST, foundational evidence for AHA/ASA Class I, Level A early aspirin within 24-48h once ICH is excluded.