The Clotbust Review

Tenecteplase in Large-Vessel Occlusion — Bridging, Extended-Window, and the 4.5-Hour Wall

· DOI: 10.3389/fneur.2025.1715049 · PMC12834819 · stroke deep-dive tenecteplase thrombectomy large-vessel-occlusion meta-analysis

Stylized illustration of thrombolytic drug action in a cerebral vessel with retriever
AI-generated editorial illustration. Generated for editorial use.

Clinical Question (PICO)

  • P: Adults with acute ischemic stroke due to large-vessel occlusion (LVO; anterior or posterior circulation), including patients destined for endovascular thrombectomy (EVT) and those without EVT access, treated within the standard (≤4.5 h) or extended (4.5–24 h, imaging-selected) windows.
  • I: Intravenous tenecteplase 0.25 mg/kg single bolus.
  • C: No intravenous thrombolysis (with or without subsequent EVT).
  • O: Disability shift on the modified Rankin Scale (mRS) at 90 days; mRS 0–1 (excellent outcome); mRS 0–2 (functional independence); symptomatic intracerebral hemorrhage (sICH); all-cause mortality. Prespecified subgroups: EVT vs no EVT; treatment window ≤4.5 h vs 4.5–24 h.

Bottom Line

Across 5 RCTs and 2,054 patients, IV tenecteplase increased the odds of a better 90-day mRS outcome (generalized OR 1.19; 95% CI 1.06–1.34) and the rate of functional independence (RR 1.14; 95% CI 1.03–1.26) versus no thrombolysis, without an increase in sICH (RR 1.46; 95% CI 0.87–2.54) or mortality (RR 1.07; 95% CI 0.84–1.36). The benefit is concentrated in two clinical niches — bridging within 4.5 h before EVT, and extended-window treatment of patients without EVT access but with salvageable tissue on CT perfusion — while bridging beyond 4.5 h in EVT-treated patients appears non-beneficial and may worsen functional independence.

Design

  • Trial type: Systematic review and meta-analysis of randomized clinical trials, with a target-trial emulation study included as one of the five evidence sources.
  • N: 2,054 patients across 5 studies; 1,000 received tenecteplase.
    • BRIDGE-TNK (2025): 550 patients, 278 tenecteplase
    • TRACE III (2024): 516 patients, 264 tenecteplase
    • TIMELESS (2024): 458 patients, 228 tenecteplase
    • CHABLIS-T II (2025): 224 patients, 111 tenecteplase
    • Target trial emulation of SWIFT-DIRECT + EXTEND-IA TNK (2025): 306 patients, 122 tenecteplase (0.25 mg/kg arm only)
  • Search strategy: MEDLINE/PubMed, Embase, Cochrane Library, Web of Science, last run November 2025; manual reference and forward-citation searches.
  • Setting: International, academic stroke centers; trials enrolled across multiple countries.
  • Analysis: Random-effects meta-analysis with generalized odds ratios (gORs) for mRS shift; risk ratios (RRs) with 95% CIs for binary outcomes; prespecified subgroups by EVT status and treatment window (≤4.5 h vs 4.5–24 h); I² reported for heterogeneity.
  • Risk of bias / sensitivity: Sensitivity analysis excluding CHABLIS-T II yielded consistent results.
  • Primary outcome: Disability shift on mRS at 90 days, plus binary mRS 0–1, mRS 0–2, sICH, mortality.
  • Protocol registration: PROSPERO CRD420251115406; PRISMA-compliant.

Population

Inclusion Criteria

  • RCTs or secondary analyses of RCTs in LVO patients (anterior or posterior circulation; intracranial ICA, M1, M2, ACA, or basilar artery).
  • Comparator: tenecteplase 0.25 mg/kg vs no thrombolysis; trials in which only a minority of comparators received an alternative thrombolytic were permitted.
  • Required stratification by EVT status (yes/no).
  • No restriction on treatment time window.

Exclusion Criteria

  • Observational studies, conference abstracts, non-English publications.
  • Trials where the comparator arm predominantly received another thrombolytic (e.g., alteplase).

Baseline Characteristics (representative — BRIDGE-TNK)

  • Anterior circulation LVO (ICA/M1/M2) in 91% (the remaining 9% were vertebrobasilar occlusions).
  • Standard imaging selection (no CT perfusion requirement); enrollment within 4.5 h of onset.

Interventions

  • Tenecteplase arm: IV tenecteplase 0.25 mg/kg, single bolus, delivered before EVT (bridging) or as primary therapy in patients without EVT access.
  • Control arm: No IV thrombolysis. Patients in both arms could still proceed to EVT if eligible; for the no-EVT extended-window subgroup, CT perfusion selection for salvageable tissue was required.

Outcomes

Primary Outcome (mRS shift at 90 days, gOR):

  • Overall tenecteplase vs no thrombolysis: gOR 1.19 (95% CI 1.06–1.34) — favors tenecteplase.
  • No significant interaction between tenecteplase benefit and EVT status (interaction p non-significant).

Subgroup: Patients undergoing EVT

  • Within 4.5 h (bridging): mRS shift gOR 1.24 (95% CI 1.03–1.50; p=0.03; I²=48%); mRS 0–2 RR 1.16 (95% CI 1.02–1.31; p=0.03; I²=0%). Favors tenecteplase.
  • Between 4.5 and 24 h: mRS shift gOR 0.95 (95% CI 0.70–1.30; p=0.76; I²=78%); mRS 0–2 RR 0.60 (95% CI 0.40–0.90; p=0.01). No benefit on disability; worse functional independence with tenecteplase.
  • sICH and mortality similar between groups (Supplementary Figure S3).

Subgroup: Patients NOT undergoing EVT, extended window (>4.5 h), CTP-selected

  • mRS 0–1 (excellent outcome) RR 1.41 (95% CI 1.12–1.78; p=0.004; I²=0%).
  • mRS 0–2 (functional independence) RR 1.25 (95% CI 1.04–1.49; p=0.01; I²=0%).
  • mRS shift gOR 1.28 (95% CI 0.98–1.67; p=0.07; I²=0%) — directional but non-significant.

Overall Binary Outcomes (tenecteplase vs no thrombolysis)

  • Excellent outcome (mRS 0–1): RR 1.21 (95% CI 1.06–1.38).
  • Functional independence (mRS 0–2): RR 1.14 (95% CI 1.03–1.26).
  • sICH: RR 1.46 (95% CI 0.87–2.54) — non-significant.
  • Mortality: RR 1.07 (95% CI 0.84–1.36) — non-significant.

Adverse Events / Safety

  • No statistically significant increase in symptomatic intracerebral hemorrhage or all-cause mortality in any subgroup.
  • Authors note that the sICH point estimate (RR 1.46) does not exclude clinically meaningful harm; the wide CI (0.87–2.54) reflects limited event counts across 5 trials.

Figures

Forest plots of overall treatment effects comparing intravenous tenecteplase wit
Figure 1. Forest plots of overall treatment effects comparing intravenous tenecteplase with no thrombolysis among patients with LVO, showing mRS shift analysis (A) , excellent functional outcome (mRS 0–1) (B) , and functional independence (mRS 0–2) (C) ..

Source: PMC PMC12834819 — fneur-16-1715049-g001.jpg. Click image to expand.

Forest plots of subgroup analyses by treatment timing among patients with LVO un
Figure 2. Forest plots of subgroup analyses by treatment timing among patients with LVO undergoing EVT, showing mRS shift analysis (A) and functional independence (mRS 0–2) (B) ..

Source: PMC PMC12834819 — fneur-16-1715049-g002.jpg. Click image to expand.

Forest plots of subgroup analyses by treatment timing among patients with LVO no
Figure 3. Forest plots of subgroup analyses by treatment timing among patients with LVO not undergoing EVT, showing mRS shift analysis (A) , excellent functional outcome (mRS 0–1) (B) , and functional independence (mRS 0–2) (C) ..

Source: PMC PMC12834819 — fneur-16-1715049-g003.jpg. Click image to expand.

Criticisms

  • Small study count (n=5). Only one to two trials contribute to each subgroup, which limits the precision of the 4.5–24 h EVT-bridging estimate (I²=78% for the shift analysis).
  • Comparator heterogeneity. The target-trial emulation aggregates SWIFT-DIRECT and EXTEND-IA TNK, mixing populations. The strict inclusion criterion — that the comparator arm received minimal alternative thrombolytic — implicitly down-weights trials like AcT (tenecteplase vs alteplase).
  • Heterogeneity in extended-window estimates. I²=78% for the late-window bridging mRS shift, driven by differences in patient mix and imaging protocols.
  • Patient-level data absent. Authors flag the lack of IPD as a barrier to granular subgroup exploration.
  • Posterior circulation caveat. Only BRIDGE-TNK included vertebrobasilar occlusions (49; ~9%); generalizability to posterior LVO is limited.
  • External validity. All trials required advanced imaging selection (CT perfusion or comparable) for late-window inclusion; results may not generalize to settings without CTP access, including many low- and middle-income systems.
  • sICH imprecision. CI crosses 1.0 (0.87–2.54), but the upper bound (2.54) cannot exclude a clinically important doubling — the safety verdict is “no signal” not “proven safe.”

Funding

The authors declared that no financial support was received for this work or its publication. The study was conducted in the absence of any commercial or financial relationships that could constitute a conflict of interest. The protocol was registered with PROSPERO (CRD420251115406). No generative AI was used in manuscript preparation (figure alt-text generated by the publisher).

The paper

  • Authors. Alkhiri A, Alamri AF, Alturki F, Alonazi RM, Nguyen TN, Al-Ajlan FS, Alhazzani A.
  • Title. Tenecteplase in large-vessel occlusion with or without thrombectomy: a meta-analysis.
  • Journal. Frontiers in Neurology.
  • Year. 2025.
  • DOI. 10.3389/fneur.2025.1715049
  • PMCID. PMC12834819
Deep Dive — click to expand

What this is

A 2,054-patient meta-analysis of five RCTs asking whether IV tenecteplase 0.25 mg/kg improves 90-day mRS outcomes in large-vessel occlusion stroke — separating patients who go on to thrombectomy from those who do not, and stratifying by the 4.5-hour treatment wall. The headline: tenecteplase is a real win when given within 4.5 h before EVT, and a real win when given in the extended window to imaging-selected patients without EVT access — but the bridging indication beyond 4.5 h looks dead, with a signal toward worse functional independence (RR 0.60). This is the most decisive read yet on where tenecteplase belongs in the LVO pathway.

1. Shadow Audit

The paper frames the sICH result as reassuring (“no significant increase,” RR 1.46 with CI 0.87–2.54). Read the CI: it crosses 1.0 but its upper bound sits at 2.54 — meaning the data are consistent with a doubling of symptomatic brain bleeds and we cannot rule it out. With only 2,054 patients and low event counts, the safety verdict is “underpowered to detect a clinically meaningful harm,” not “demonstrated safe.” The paper also pools across trials with different sICH definitions (ECASS-II vs SITS-MOST vs NINDS-style) without harmonizing them — a methodological choice that quietly attenuates detection of bleeding risk. The authors acknowledge the bleeding imprecision in the discussion but the abstract’s “without an increased risk” reads more confidently than the evidence supports.

2. Inversion Engine

For the headline result (gOR 1.19 favoring tenecteplase) to flip, you would need the bulk of the mRS-shift advantage to be attributable to either (a) selection bias in the trials — sicker patients randomized away from tenecteplase, (b) unmeasured co-intervention differences (e.g., faster EVT door-to-puncture times in the tenecteplase arms because pre-EVT lysis unblocks the operator), or (c) a single outlier trial like CHABLIS-T II driving the win. The sensitivity analysis excluding CHABLIS-T II yielded consistent results, so (c) is partially defused. (a) is theoretically possible but the RCT design controls it. (b) is the most plausible inversion — and it would mean tenecteplase isn’t doing the work; the workflow is. Quantitatively: if the entire effect (gOR 1.19, log-scale shift of ~0.17) disappeared, the lower 95% CI bound (1.06) would still leave a tiny positive effect — so you can’t fully invert without a structural workflow asymmetry.

3. Second-Order Catalyst

The first protocol that moves is the rural and spoke-hospital drip-and-ship pathway: tenecteplase 0.25 mg/kg becomes the default for any LVO patient who can be at a CT within 4.5 h, even if the comprehensive stroke center is 60 minutes away by air. Comprehensive centers update their door-in-door-out targets to assume tenecteplase is already on board. The first institutional mover is going to be a regional telestroke network — the patient subgroup that moves first is the late-presenting rural LVO without thrombectomy access, who now has a proven pharmacological option selected by CTP. In tertiary centers with mature EVT pathways, the change is subtler: the 4.5–24 h bridge becomes a hard contraindication, not a soft preference.

4. Asymmetric Leverage

The asymmetric payoff sits in the extended-window no-EVT subgroup: RR 1.41 for excellent outcome (mRS 0–1), absolute risk increase roughly 13 percentage points based on baseline rates, with no detectable mortality cost. That denominator is small in absolute terms (the no-EVT population is a minority of LVO), but per patient treated it is the highest yield win in the entire meta-analysis. The opposite asymmetry is the 4.5–24 h bridging result: a small denominator (a few hundred patients across TIMELESS + CHABLIS-T II + TRACE III), high heterogeneity (I²=78%), and a directional signal toward harm (RR 0.60 for mRS 0–2) — i.e., small effect, large downside tail.

5. Paradigm Destroyer

The reflex this kills: giving tenecteplase to a confirmed-LVO patient arriving between 4.5 and 24 h on the theory that “something is better than nothing before the thrombectomy.” The 2024–2025 evidence base — TIMELESS, CHABLIS-T II, TRACE III — converges on the same answer: late-window bridging does not improve mRS shift and may worsen functional independence. The two-sentence protocol change for tomorrow morning: (1) if your LVO patient is inside 4.5 h, give tenecteplase and ship for EVT; (2) if your LVO patient is 4.5–24 h, skip the bridge and go straight to EVT unless they have no EVT access — in which case, CTP-select and give tenecteplase as primary therapy. The middle option — late bridge — is no longer a defensible default.

MVP — Minimum Viable Proof

A single demonstrable claim sufficient to update practice: in LVO patients treated with IV tenecteplase before EVT within 4.5 h, the rate of functional independence at 90 days is 16% higher than with EVT alone (RR 1.16; 95% CI 1.02–1.31; p=0.03). If true — and BRIDGE-TNK plus the EXTEND-IA TNK-derived emulation independently support it — then the drip-and-ship pathway for early-window LVO should default to tenecteplase, not alteplase, and certainly not “no lytic.”

Best Combination

Stack this with the prior literature: (a) EXTEND-IA TNK (Campbell et al., 2018) showed tenecteplase achieves higher early reperfusion than alteplase in patients going to EVT; (b) the original NOAC-IVT, AcT, and TASTE trials established tenecteplase non-inferiority to alteplase in non-LVO populations; (c) IRIS (2024) showed alteplase bridging in EVT was non-beneficial. The Alkhiri meta-analysis is the tenecteplase-mirror of IRIS — and it lands the opposite way inside the 4.5 h window, suggesting that tenecteplase’s faster recanalization kinetics buy enough time for the bridge to matter before alteplase’s slower lytic would. Read together: tenecteplase has both a broader eligibility (extended window without EVT, imaging-selected) and a tighter efficacy signal inside the early-window bridge than alteplase.

Overvalue Warning

Two specific overreads to flag. First: the headline gOR 1.19 sounds like a 19% improvement, but on the generalized odds ratio scale that is a much smaller absolute effect on the probability of mRS 0–2 than the number suggests — read the mRS 0–2 RR (1.14) for the more clinically interpretable effect size. Second: the “no sICH increase” framing will be cited as proof of safety; the CI (0.87–2.54) does not support that read. Treat the safety verdict as “no signal in this dataset,” not “demonstrated safe,” and remember that an RR of 1.46 for brain bleeds in a small pooled population is exactly the kind of signal that becomes statistically significant the moment a sixth large trial publishes.

Educational use notice. The Clotbust Review is dedicated to medical education. Figures from published articles are embedded for teaching purposes, with attribution to the original publication. We do not claim copyright over any embedded figure; all rights remain with the original publisher and authors under the article's published license. If you are a rights holder and would like a figure removed, please contact the editorial team. See our editorial notice for the full policy.