Argatroban with antiplatelets did not improve stroke recovery, and bled more in trials

Design

Systematic review and meta-analysis of 17 studies (randomised and observational)

Population

About 70,000 patients with acute ischaemic stroke

Primary outcome

mRS 0–2 and composite haemorrhage

Effect

mRS 0–2 RR 1.03 (95% CI 0.97–1.10); composite haemorrhage RR 1.10 (0.98–1.23)

This meta-analysis pooled 17 studies, randomised and observational, comparing argatroban added to antiplatelet therapy with antiplatelet therapy alone in acute ischaemic stroke, covering about 70,000 patients.

The main functional outcome, a modified Rankin score of 0 to 2, did not differ (RR 1.03). Early neurological deterioration was roughly halved (RR 0.48), and an excellent outcome (mRS 0–1) was slightly more frequent (RR 1.22). Across all studies bleeding did not differ, but when the analysis was restricted to randomised trials, composite and intracranial haemorrhage were significantly higher with argatroban.

Argatroban is used in parts of Asia for progressing non-cardioembolic stroke. The randomised bleeding signal matters more than the pooled reassurance, because observational studies are prone to giving the drug to lower-risk patients.

The certainty of evidence ranged from very low to high across outcomes. This does not support routine use, and any benefit on early deterioration needs weighing against bleeding.

• Do not add argatroban routinely to antiplatelet therapy in acute ischaemic stroke on this evidence.
• If argatroban is considered for progressing stroke, weigh the higher intracranial bleeding seen in randomised trials.
• Follow local stroke protocols for early deterioration, including repeat imaging and blood pressure review.
• Give more weight to randomised results than to pooled observational data when the two disagree.

Why it matters

An apparently safe adjunct turns out to bleed more when only randomised data are examined.

Don't overread it

Fewer early deteriorations did not translate into better function at follow-up.

The statistics, in plain English

A risk ratio of 1.03 (0.97 to 1.10) for good functional outcome includes no difference. The early deterioration result (RR 0.48) looks large, but pooled data mixing observational and randomised studies can exaggerate benefit. When randomised and observational results disagree on harm, the randomised result is usually more trustworthy.

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.