Tenecteplase before thrombectomy helped only where the infarct was already large

Design

post hoc subgroup analysis of a randomised trial, with treatment-by-ASPECTS interaction testing

Population

550 patients with acute large-vessel occlusion within 4.5 hours, across China, May 2022 to September 2024

Primary outcome

90-day functional independence (mRS 0–2), with 48-hour symptomatic haemorrhage and 90-day mortality

Effect

ASPECTS <8: aRR 1.67 (95% CI 1.18–2.35); ASPECTS 8–10: aRR 1.00 (0.84–1.17), interaction p = 0.007

BRIDGE-TNK randomised patients with large-vessel occlusion within 4.5 hours of last known well to intravenous tenecteplase before endovascular thrombectomy or to thrombectomy alone. This post hoc analysis split its 550 patients by baseline ASPECTS: 241 (43.8%) scored under 8 and 309 scored 8 to 10, with median age 69 and 70 respectively.

The interaction was strong. Functional independence at 90 days — modified Rankin 0 to 2 — favoured tenecteplase in the ASPECTS under 8 group (adjusted risk ratio 1.67, 95% CI 1.18 to 2.35) but was flat in the ASPECTS 8 to 10 group (1.00, 0.84 to 1.17), interaction p = 0.007. Symptomatic intracranial haemorrhage at 48 hours did not differ significantly in either stratum, though the numbers ran the other way in the higher-ASPECTS group (7.5% with tenecteplase against 2.8% without). Ninety-day mortality was comparable where ASPECTS was under 8 but numerically higher with tenecteplase where it was 8 to 10 (aRR 1.89, 0.99 to 3.61, interaction p = 0.04).

This inverts the intuition. The patients with more preserved tissue — the ones thrombolysis is usually assumed to protect — got no functional benefit and a possible mortality signal, while the benefit sat with those whose infarct was already extensive. The authors are explicit that this is exploratory and needs prospective confirmation, and it should not change bridging practice on its own. But it does mean a bridging decision in a patient with an excellent ASPECTS is less obviously right than it has been treated as.

• Record baseline ASPECTS before bridging, and record who read it
• Do not change bridging protocol on this analysis — it is post hoc and needs prospective data
• Treat a high ASPECTS as a reason to discuss the decision, not as reassurance that thrombolysis is harmless
• Watch for prospective trials stratifying by ischaemic extent; the question is now formulated

Why it matters

It challenges the assumption that bridging thrombolysis is the safe addition in patients with the most salvageable brain.

Don't overread it

This is an exploratory post hoc analysis of a single trial conducted in China — it is not a reason to withhold thrombolysis.

The statistics, in plain English

The interaction p-value of 0.007 says the treatment effect genuinely differed between the two strata rather than the subgroups differing by chance — which is the thing subgroup analyses usually cannot claim. The mortality hazard of 1.89 with an interval of 0.99 to 3.61 just touches 1.0, so it is a signal rather than a finding. Post hoc subgroup analysis remains hypothesis-generating whatever the p-value, because the question was formed after seeing the data.

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.