Thrombolysis from 4.5 to 24 hours: better function, and two and a half times the symptomatic haemorrhage

Design

systematic review and random-effects meta-analysis of 14 randomised trials

Population

4,944 patients with acute ischaemic stroke, 2,492 thrombolysed 4.5-24 hours after onset or last known well

Primary outcome

modified Rankin 0-1 at 3 months and symptomatic intracranial haemorrhage

Effect

excellent outcome odds ratio 1.43 (95% CI 1.25-1.63); recanalisation 3.28 (2.09-5.16); symptomatic haemorrhage 2.51 (1.47-4.28); mortality 1.21 (0.95-1.53)

Fourteen randomised trials, 4,944 patients, comparing intravenous thrombolysis with alteplase or tenecteplase given between 4.5 and 24 hours after onset or last known well in imaging-selected patients against standard care. Mean age was 69.8, median NIHSS 9, and 12.3% had preplanned thrombectomy.

Extended-window thrombolysis raised the odds of an excellent outcome at three months, modified Rankin 0 to 1, to 1.43 (95% CI 1.25 to 1.63), of a good outcome to 1.25 (1.11 to 1.40), and of recanalisation to 3.28 (2.09 to 5.16). Three-month mortality did not differ (1.21, 0.95 to 1.53). Symptomatic intracranial haemorrhage rose: odds ratio 2.51 (1.47 to 4.28). Excluding patients who received thrombectomy did not change the picture. Tenecteplase and alteplase were similarly effective, and the haemorrhage signal was numerically smaller with tenecteplase (1.96, 1.06 to 3.64) than alteplase (5.29, 1.80 to 15.57), though the subgroup difference was not significant at p=0.11.

The consequence for a stroke service is that the clock is no longer the gate; the scanner is. A patient arriving at eight or fourteen hours with a favourable perfusion or mismatch profile is a candidate, and the conversation with the family is a real trade - a better chance of independence against a clearly higher chance of a bleed, with no measurable difference in death. Where advanced imaging is not available, the extended window is not available either, and that, rather than the evidence, is what limits it across most of India.

• Set up the pathway so a late-presenting patient goes to advanced imaging rather than being excluded on time
• Quote both numbers to the family: better function at OR 1.43, symptomatic haemorrhage at OR 2.51
• Where both agents are available, tenecteplase is at least as effective and the bleeding signal was numerically smaller
• Audit which late patients are getting imaged - the gate usually fails at triage, not at the decision
• Be explicit in referral discussions about whether perfusion imaging is available at the receiving centre

Why it matters

It moves the decision to thrombolyse from the clock to the scan, which changes who gets triaged for imaging rather than only who gets the drug.

Don't overread it

This pools imaging-selected patients - it does not support thrombolysing a late presenter without advanced imaging.

The statistics, in plain English

An odds ratio of 1.43 for excellent outcome and 2.51 for symptomatic haemorrhage describe events of very different frequency, so the ratios are not directly comparable: functional independence is common and bleeding is rare, meaning the absolute gain is larger than the absolute harm even though the harm ratio looks bigger. The mortality interval, 0.95 to 1.53, includes both a small benefit and a meaningful increase - the trials cannot exclude either. The tenecteplase-versus-alteplase comparison is a subgroup analysis with p=0.11 for the difference: suggestive, and not a basis for choosing on bleeding risk alone.

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.