Tenecteplase, cognition and rescue stenting: Biplane Weekly No. 3

ATTENTION LATE asked whether intravenous tenecteplase (0.25 mg/kg) given before thrombectomy helps patients with basilar artery occlusion who arrive 4.5 to 24 hours after onset. The open-label, blinded-endpoint trial randomised 332 patients with an NIHSS of 10 or more at 40 Chinese centres, all admitted directly to thrombectomy-capable hospitals.

Functional independence (mRS 0–2) at 90 days was reached by 50 of 165 patients given tenecteplase and 50 of 164 treated with thrombectomy alone (adjusted rate ratio 0.92, 95% CI 0.67–1.25). Symptomatic intracranial haemorrhage occurred in 5% versus 4%, and 90-day mortality was 40% versus 43%.

ATTENTION and BAOCHE established thrombectomy itself for basilar occlusion; this trial gives no reason to add late intravenous lysis to it in patients who arrive directly. It does not address drip-and-ship pathways, where the question is whether lysis at the referring hospital buys anything before transfer.

Cognitive decline is often offered as a reason to treat asymptomatic carotid stenosis. CREST-2 tested it directly. In two parallel trials of patients with asymptomatic stenosis of 70% or more, 2165 of 2366 participants underwent blinded telephone cognitive testing: 1078 in the stenting trial and 1087 in the endarterectomy trial, each comparing revascularisation plus intensive medical management with intensive medical management alone.

At four years, the adjusted composite z score of five tests did not differ: 0.22 versus 0.18 in the stenting trial (difference 0.03, 95% CI −0.09 to 0.15) and 0.25 versus 0.23 in the endarterectomy trial (difference 0.02, −0.09 to 0.14). Both intervals exclude the prespecified minimal clinically important difference of 0.50 SD, and no individual test differed. Mean follow-up was 2.24 years, so the four-year estimates rest on fewer patients.

The CREST-H substudy, which tested whether any cognitive benefit varied with baseline haemodynamic impairment on perfusion imaging, found no differential effect over one year. The decision therefore returns to CREST-2's primary result, in which stenting had a lower four-year rate of peri-procedural stroke or death or later ipsilateral stroke than intensive medical management (2.8% versus 6.0%) and endarterectomy did not differ significantly (3.7% versus 5.3%); these rates come from all 2485 randomised patients, treated under a protocolised intensive medical regimen.

Many European patients with large-vessel occlusion are first imaged at a primary stroke centre, and a large core on that scan can end the transfer conversation. This post hoc analysis of the LASTE trial examined the 157 of the 324 enrolled patients whose ASPECTS of 0–5 was established at a primary centre before transfer, within 6.5 hours of last seen well: 77 had thrombectomy plus medical management and 80 medical management alone. Median age was 73, median NIHSS 20, and 56% had ASPECTS 0–2 on the referring scan. The analysis is post hoc and was appraised from the abstract.

Thrombectomy shifted the 90-day mRS distribution favourably (generalised OR 1.71, 95% CI 1.21–2.41), with no heterogeneity across ASPECTS strata. Mortality (RR 0.73, 0.50–1.05) and symptomatic haemorrhage (RR 1.85, 0.64–5.26) did not differ significantly.

The subgroup result is consistent with the main LASTE trial. It suggests that an ASPECTS of 0–5 on the referring scan, on its own, need not end the transfer conversation within 6.5 hours; it says nothing about patients who were never transferred or enrolled.

A new meta-analysis in the Journal of Neurology pools six randomised trials of coiling against clipping for ruptured aneurysms: Koivisto (2000), ISAT (Molyneux, 2005), Li (2012), BRAT (McDougall, 2012), Wadd (2015) and ISAT-2 (Darsaut, 2025).

At one year, mRS 0–2 was reached by 1083 of 1431 coiled and 1084 of 1613 clipped patients (OR 1.60, 95% CI 1.36–1.88; I² 0%). The wider mRS 0–3 threshold did not differ (OR 1.19, 0.96–1.48; three trials), nor did mortality (OR 0.89, 0.69–1.14). Complete occlusion at 12 months was less frequent after coiling: 701 of 1157 versus 589 of 732 (OR 0.35, 0.28–0.44; four trials). In the two trials reporting middle cerebral artery aneurysms, clipping achieved complete occlusion more often (OR 5.21, 1.57–17.30).

The functional result is largely ISAT's, which carries 67.8% of the weight, and the trials span more than two decades of device change; none tested stent-assisted, flow-diverting or intrasaccular techniques. Blinding of participants and outcome assessors was rated high risk in every trial.

The paper quantifies the trade-off clinicians already weigh: better one-year function after coiling, more complete occlusion after clipping, with the middle cerebral artery the clearest case for clipping on occlusion grounds.

A meta-analysis in Neurological Research pools six randomised trials (2061 patients) of thrombectomy against best medical therapy for distal medium vessel occlusion, the territory in which DISTAL, ESCAPE-MeVO and DISCOUNT were neutral. It was appraised from the abstract.

At 90 days, thrombectomy did not increase excellent outcome (mRS 0–1: RR 1.19, 95% CI 0.89–1.59) or functional independence (mRS 0–2: RR 1.07, 0.89–1.30), and mortality did not differ (RR 1.12, 0.73–1.71). The point estimate for symptomatic haemorrhage was more than doubled (RR 2.31, 0.98–5.46), short of significance and sensitive to the exclusion of one study.

A separate meta-analysis in Cerebrovascular Diseases, pooling three randomised trials (1233 patients), likewise found no gain in mRS 0–1 (RR 0.95, 0.82–1.10) but a significant excess of symptomatic haemorrhage (OR 2.38, 1.35–4.20). Routine thrombectomy for these occlusions remains unsupported. A further randomised test is SAIL-MEVO, newly registered by Vall d'Hebron, with 420 patients planned and functional independence at 90 days as its primary endpoint.

When thrombectomy fails to reperfuse a middle cerebral artery occlusion, the choice is often between intracranial stenting and stopping. The German Stroke Registry–Endovascular Treatment, analysed retrospectively from prospectively collected data, identified 352 stented patients (2.4%) among 14,714 eligible patients treated from 2015 to 2023. It propensity-matched 218 of them to 218 patients with failed reperfusion (mTICI below 2b) who had no further endovascular treatment.

Stenting was associated with a shift towards lower 90-day mRS (adjusted common OR 0.41, 95% CI 0.29–0.59, where values below 1 favour stenting; median mRS 3 versus 5). Mortality was 26% versus 37% and malignant MCA infarction 1% versus 8%, while symptomatic haemorrhage did not differ (4% versus 5%).

The comparator is a group in whom operators chose to stop. Stent type and antiplatelet regimen were not uniformly captured and the reason for stenting not at all, and 38% of stented patients were excluded for missing covariates. The only randomised comparison the authors discuss, ANGEL-REBOOT, showed no 90-day benefit and more complications with rescue stenting, although extended follow-up reported better one-year outcomes. The registry cannot separate the effect of the stent from the selection of patients with likely underlying atherosclerosis, but in this matched cohort stenting carried no excess of symptomatic haemorrhage.

An editorial in Neurochirurgie by Comby, Benkerroum, Darsaut and Raymond examines the composite primary endpoints of five randomised trials of middle meningeal artery embolisation for chronic subdural haematoma. Three of the five reported that embolisation reduced haematoma progression or recurrence; none showed an effect on other clinical outcomes such as neurological function. The authors argue that the composite benefit may be artefactual, and a second editorial in the same issue asks whether the results are too good to be true.

Registry data published the same month show how hard the radiological gain is to pin down. In the MESH registry (679 patients, liquid embolic), multi-pedicle embolisation was associated with greater haematoma thinning at one month than single-pedicle embolisation (5.7 versus 3.3 mm), but not at three months (8.2 versus 8.4 mm), and recurrence and rescue surgery did not differ; a higher rate of mRS 0–2 (53.5% versus 43.0%) is open to confounding by indication.

As published, the trials support a claim about progression or recurrence as each trial defined it, not about neurological function, and the editorial questions whether even that composite gain is robust.