Branch atheromatous disease stroke is the one that gets worse in front of you. Early neurological deterioration is common, and the absence of any proven treatment has driven widespread off-protocol use of intensive antiplatelet therapy. STRATEGY tested the most popular version of that reflex.
This double-blind placebo-controlled trial across 38 Chinese hospitals randomised 970 patients with MRI-confirmed branch atheromatous disease stroke within 48 hours of onset to intravenous tirofiban or placebo — 0.4 micrograms per kilogram per minute for 30 minutes, then 0.1 for 24 hours — on top of aspirin 300 mg loading and 100 mg daily to day 90. Median age was 63.
The primary endpoint of early neurological deterioration within 7 days or new stroke within 90 days occurred in 79 patients (17.1%) on tirofiban and 89 (19.6%) on placebo, hazard ratio 0.88 (95% CI 0.65 to 1.19, p=0.39). Moderate or severe bleeding occurred in one patient on tirofiban and none on placebo.
So the drug is safe and it does not work. That combination is awkward, because a safe ineffective drug is the hardest kind to stop using — there is no adverse event to point at, and the patient who does well is remembered. The finding to hold onto is the placebo arm's event rate: nearly one in five deteriorated or had another stroke despite aspirin. That is the size of the unmet need, and it is not met by adding a glycoprotein IIb/IIIa inhibitor.
- Do not add intravenous tirofiban to aspirin for branch atheromatous disease stroke; it did not reduce deterioration or recurrence
- The absence of bleeding harm is not a reason to continue using it — an ineffective drug still costs money and monitoring
- Expect roughly one in five of these patients to deteriorate or have another stroke within 90 days on aspirin alone
- Continue standard secondary prevention: antiplatelet therapy, blood pressure control, lipid lowering and glycaemic control
- Note the trial required MRI confirmation of branch atheromatous disease, so it does not speak to undifferentiated lacunar stroke
The statistics, in plain English
A hazard ratio of 0.88 with an interval from 0.65 to 1.19 crosses 1.0, so no effect can be claimed, but the interval's lower end leaves room for a 35% reduction that the trial was too small to detect. With 168 events in 970 patients this is a moderately powered null rather than a definitive one. Read the absolute figures: 17.1% versus 19.6%, a 2.5 percentage point difference that could easily be chance. The safety result — one bleeding event in 486 patients — is reassuring but is also why an ineffective drug of this kind persists in practice.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for neurology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free