Factor XIa inhibitors cut recurrent stroke by a quarter with no measurable bleeding cost

Design

systematic review and random-effects meta-analysis of three randomised placebo-controlled trials

Population

14,239 participants with acute noncardioembolic ischaemic stroke or TIA (6,927 assigned to a factor XIa inhibitor)

Primary outcome

first stroke of any type during follow-up; first major bleeding event as primary safety outcome

Effect

any stroke RR 0.75 (95% CI 0.66-0.84); ischaemic stroke RR 0.74 (0.66-0.84); composite cardiovascular events RR 0.83 (0.75-0.92); major bleeding RR 1.12 (0.87-1.44)

Three randomised trials comparing a factor XIa inhibitor with placebo, added to standard antiplatelet therapy after acute noncardioembolic ischaemic stroke or transient ischaemic attack, were pooled: 14,239 participants, of whom 6,927 received the drug. Any stroke fell with a risk ratio of 0.75, ischaemic stroke 0.74, and composite cardiovascular events 0.83.

The safety result is the reason this class exists. Major bleeding showed a risk ratio of 1.12 with a confidence interval spanning 1.0, and there was no increase in haemorrhagic stroke, intracranial haemorrhage, any bleeding or all-cause mortality. The pharmacological premise — that inhibiting factor XIa interrupts pathological thrombosis while leaving haemostasis largely intact — is what every previous attempt to intensify antithrombotic therapy after stroke has failed to deliver. This is the first pooled evidence that it may hold.

What it changes is the framing of a decision clinicians currently have no good answer to. A patient who has a recurrent ischaemic stroke on adequate antiplatelet therapy is common, and the options have been to add a second antiplatelet, accept the bleeding risk of anticoagulation, or do nothing different. A third possibility is now credible.

It is not yet a prescribing decision. No factor XIa inhibitor is available for this indication in India, and the pooled estimate rests overwhelmingly on one trial.

• Do not change antithrombotic practice on this; no agent in the class is licensed for secondary stroke prevention here
• Expect the question from patients with recurrent stroke on antiplatelet therapy who have read about the class
• Continue to optimise what is proven — blood pressure, lipids, glycaemia and the antiplatelet already prescribed
• Note which patients on your list have had recurrence despite adherence; that is the group these trials targeted
• Watch for the phase 3 readouts individually rather than relying on the pooled figure

The statistics, in plain English

A risk ratio of 0.75 with a 95% confidence interval of 0.66 to 0.84 means a quarter fewer strokes, and the interval sits entirely below 1.0, so the reduction is unlikely to be chance. The bleeding result, 1.12 with an interval of 0.87 to 1.44, crosses 1.0 — that is reassurance rather than proof of safety, and the upper bound leaves room for up to a 44% relative increase in major bleeding that this sample size could not detect. The authors state plainly that the estimates are largely driven by a single phase 3 trial of 12,327 participants; pooling three trials when one supplies 87% of the participants does not give three trials' worth of independent confirmation.

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.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.