The analysis, led by Wenkui Li and colleagues at Chongqing University Three Gorges Hospital in China, pooled data from fifteen randomized clinical trials that tested reperfusion treatments administered between 4.5 and 24 hours after stroke onset or the last time a patient was known to be well. This extended window matters enormously in clinical practice. Roughly one in five ischemic strokes is discovered only when the patient wakes up or is found after the standard 4.5-hour thrombolysis window has closed, leaving clinicians with difficult decisions about whether aggressive intervention can still salvage brain tissue that is starved of blood but not yet irreversibly damaged.

The researchers took particular care to separate patients with radiologically confirmed intracranial large-vessel occlusion, or LVO, from broader mixed populations of stroke patients. Large-vessel occlusion refers to blockages of the major arteries supplying the brain, such as the middle cerebral artery or the basilar artery, and these blockages are the targets that mechanical thrombectomy is designed to clear. Eight of the fifteen trials contributed data to the primary analysis network restricted to confirmed LVO or basilar artery occlusion. The team also ran a sensitivity analysis limited to anterior-circulation occlusions and an exploratory analysis including all available trials, and the direction of findings held consistent across all three networks.

The statistical engine behind the comparison was a Bayesian random-effects consistency model, a framework that allows indirect comparisons between treatments that were never tested against each other in the same trial. Rather than producing a single p-value, Bayesian methods generate a probability distribution around each effect estimate, expressed here as odds ratios with 95 percent credible intervals. The primary outcome was functional independence at 90 days, defined as a score of 0 to 2 on the modified Rankin Scale, the standard measure that captures whether a stroke survivor can live and function without significant assistance.

Against placebo or best medical care, endovascular thrombectomy produced the most consistent and robust evidence of benefit in the confirmed LVO population. The odds of achieving functional independence were nearly tripled, with an odds ratio of 2.93 and a credible interval of 1.40 to 6.53. Excellent functional outcome, meaning a modified Rankin Scale score of 0 to 1, was also significantly more likely with an odds ratio of 2.22. Perhaps most strikingly, thrombectomy reduced 90-day mortality by roughly 29 percent, with an odds ratio of 0.71, and increased the rate of trial-defined recanalization or reperfusion at approximately 24 hours more than fivefold, with an odds ratio of 5.27.

These gains came with a recognized trade-off. Symptomatic intracranial hemorrhage, the most feared complication of reperfusion therapy, was significantly more frequent after thrombectomy, with an odds ratio of 2.65 and a credible interval of 1.28 to 5.81. This elevated bleeding risk is a known consequence of restoring blood flow to tissue that has been deprived for hours, and it underscores why careful imaging selection of patients remains central to extended-window treatment. The benefit-risk calculus still favored intervention decisively, but the analysis makes explicit that the procedure is not without hazard.

The question of whether adding intravenous tenecteplase before thrombectomy, the so-called bridging strategy, improves outcomes received a more equivocal answer. Compared with thrombectomy alone, the combination of tenecteplase plus thrombectomy showed numerically favorable point estimates for functional independence, with an odds ratio of 1.10, and for excellent functional outcome, with an odds ratio of 1.16. However, the credible intervals were wide and crossed unity, spanning 0.34 to 3.52 and 0.52 to 2.64 respectively, meaning the differences were statistically inconclusive. Direct pairwise analyses of the two trials that compared these strategies head-to-head reached similarly uncertain conclusions. The incremental benefit of bridging tenecteplase, the authors conclude, remains unproven.

Tenecteplase given alone, without subsequent thrombectomy, did not show clear differences from placebo or best medical care in the confirmed LVO network. This finding is notable given the recent wave of trials that have established tenecteplase as a potent thrombolytic agent. Drugs such as the ones tested in studies including TNK-PLUS, CHABLIS-T II, and the tenecteplase trials published in the New England Journal of Medicine and JAMA have reshaped acute stroke care in earlier time windows, but the present analysis suggests that in the extended window with a confirmed large-vessel blockage, pharmacological clot dissolution alone is not a substitute for mechanical removal.

The technical sophistication of the analysis deserves emphasis. The researchers searched PubMed, Embase, Web of Science, and the Cochrane Library, assessed risk of bias using the revised Cochrane RoB 2 tool, and evaluated confidence in the network estimates with the CINeMA framework, a structured approach designed specifically for grading the credibility of network meta-analysis findings. All data were derived from published trial reports, and the authors have made their extracted arm-level dataset, analysis code, and CINeMA input files available upon reasonable request, a level of transparency that strengthens the reproducibility of the work.

For clinicians, the practical message is that among imaging-selected patients with confirmed intracranial large-vessel or basilar artery occlusion treated 4.5 to 24 hours after onset, endovascular thrombectomy carries the most consistent evidence of benefit, and the routine addition of tenecteplase before the procedure cannot yet be justified as superior. For the growing population of patients whose strokes are discovered late, that clarity is more than academic. It means the fastest route to the angiography suite, not the pharmacy, offers the best documented chance of walking out of the hospital independently.

Subject of Research: Comparative effectiveness of thrombolytics and thrombectomy in extended-window acute ischemic stroke with large-vessel occlusion

Article Title: Tenecteplase, alteplase, and endovascular thrombectomy in extended-window acute ischemic stroke with intracranial large-vessel occlusion: a Bayesian network meta-analysis

Article References: Li, W., Wu, C., Deng, R., Zhou, J., Li, L., Liu, Y., & Chen, S. (2026). Tenecteplase, alteplase, and endovascular thrombectomy in extended-window acute ischemic stroke with intracranial large-vessel occlusion: a Bayesian network meta-analysis. Journal of Neurology, 273(11), Article 666. https://doi.org/10.1007/s00415-026-14201-5

Image Credits: AI Generated

DOI: 10.1007/s00415-026-14201-5

Keywords: stroke, endovascular thrombectomy, tenecteplase, alteplase, network meta-analysis, neurology, large-vessel occlusion, endovascular, thrombectomy, extended-window, acute, ischemic