← Back
NeuroTrials.ai
Neurology Clinical Trial Database

ETIS-TETRIS

IV Tenecteplase Before Thrombectomy Compared With Thrombectomy Alone in Patients With Stroke Due to a Large Vessel Occlusion

Year of Publication: 2026

Authors: Gaspard Gerschenfeld, Bertrand Lapergue, Pierre Seners, ..., Sonia Alamowitch

Journal: Neurology

Citation: Neurology 2026;106:e214702

Link: https://doi.org/10.1212/WNL.0000000000214702


Clinical Question

Is IV tenecteplase plus endovascular thrombectomy (TNK + EVT) associated with better 3-month functional outcomes compared with EVT alone in patients with anterior circulation LVO stroke treated within 4.5 hours, and does any benefit vary with expected onset-to-thrombolysis time or admission center type?

Bottom Line

In this large real-world pooled registry analysis, TNK + EVT was associated with significantly better 3-month functional outcomes compared with EVT alone, without an increase in hemorrhagic complications. The benefit was not time-dependent and did not differ by admission center type (primary vs. comprehensive stroke center), supporting routine tenecteplase use before thrombectomy in the early time window.

Major Points

  • Retrospective pooled analysis of two nationwide prospective French registries: TETRIS (TNK + EVT) and ETIS (EVT alone), covering January 2015 to March 2024
  • 1,890 patients analyzed after propensity score overlap weighting (PSOW) to balance baseline confounders
  • TNK + EVT was associated with better 3-month ordinal mRS shift (OR 1.53, 95% CI 1.29-1.82, p < 0.001) and functional independence (mRS <=2: 51.0% vs. 40.4%, OR 1.50, p < 0.001)
  • No significant difference in sICH (4.5% vs. 4.5%, p = 0.61) or parenchymal hematoma (12.1% vs. 10.3%, p = 0.12)
  • No significant treatment effect modification across onset-to-thrombolysis time (p-interaction = 0.11), contrasting with alteplase data from the IRIS meta-analysis
  • No significant interaction between admission center type (PSC vs. CSC) and treatment effect (p-interaction = 0.12), though numerically higher OR in PSC-admitted patients (1.75 vs. 1.29)
  • Lower EVT rate in TNK + EVT group (85.5% vs. 96.2%) likely reflecting pre-EVT recanalization attributable to tenecteplase
  • Class II evidence per AAN classification

Design

Study Type: Retrospective pooled analysis of two prospective multicenter nationwide real-world registries

Randomization:

Blinding: No blinding; outcomes assessed unblinded in real-world conditions

Enrollment Period: January 2015 to March 2024 (ETIS); January 2015 to December 2022 (TETRIS)

Follow-up Duration: 3 months

Countries: France

Sample Size: 1890

Analysis: Propensity score overlap weighting (PSOW) with multiple imputation for missing data; propensity score-weighted ordinal and binary logistic regression; Rubin rules for pooling across imputed datasets; restricted cubic splines for nonlinear onset-to-thrombolysis time analysis; product terms for subgroup interaction testing; SAS 9.4


Inclusion Criteria

  • Adult patients with anterior circulation LVO stroke (intracranial ICA or M1/M2 MCA occlusion, with or without tandem cervical ICA occlusion)
  • Treatment with IV tenecteplase 0.25 mg/kg (max 25 mg) intended for EVT (TETRIS), or intended for EVT alone without contraindication to IVT except infarct size (ETIS)
  • Expected onset-to-thrombolysis time within 4.5 hours of known symptom onset
  • Not on anticoagulants prior to stroke

Exclusion Criteria

  • Anticoagulant use before stroke
  • Missing symptom onset-to-baseline brain imaging time
  • Isolated cervical ICA occlusion
  • Distal intracranial occlusion
  • Unknown symptom onset
  • Expected onset-to-thrombolysis time >270 minutes (TETRIS only)
  • Cerebral/systemic bleeding risk (ETIS only)
  • Other IVT contraindications (ETIS only)
  • Missing intracranial occlusion data
  • Multiple intracranial occlusions (ETIS only)

Arms

FieldTNK + EVTControl
InterventionIV tenecteplase 0.25 mg/kg (maximum 25 mg) followed by intended endovascular thrombectomy; drawn from TETRIS registryEndovascular thrombectomy alone without prior IV thrombolysis; drawn from ETIS registry
DurationSingle IV bolus at stroke presentation; EVT performed subsequentlyEVT performed at presentation

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
3-month functional outcome assessed by ordinal modified Rankin Scale (mRS) shift analysis - improvement of at least 1 point on the mRSPrimaryMedian mRS 3 (IQR 1-5)Median mRS 2 (IQR 1-4)1.53<0.001
3-month functional independence (mRS <=2)Secondary441/1,092 (40.4%)407/798 (51.0%)1.5<0.001
Final favorable recanalization (mTICI 2b-3)Secondary925/1,059 (87.4%)672/778 (86.4%)1.040.78
Parenchymal hematomaAdverse92/895 (10.3%)95/785 (12.1%)1.290.12
Symptomatic intracerebral hemorrhage (modified SITS-MOST)Adverse40/895 (4.5%)35/785 (4.5%)1.130.61

Subgroup Analysis

Benefit of TNK + EVT was consistent regardless of admission to a primary stroke center (PSC) vs. comprehensive stroke center (CSC) (p-interaction = 0.12); numerically higher OR in PSC-admitted patients (OR 1.75, 95% CI 1.39-2.20) vs. CSC-admitted patients (OR 1.29, 95% CI 0.98-1.71). No significant treatment effect modification across expected onset-to-thrombolysis time (p-interaction = 0.11 for ordinal mRS; 0.27 for mRS <=2); benefit remained statistically significant up to an expected onset-to-thrombolysis time of approximately 190 minutes. Sensitivity analysis excluding large infarct core (ASPECTS <=5) and clearly established infarcts confirmed the primary result (OR 1.59, 95% CI 1.31-1.92, p < 0.001); in this sensitivity analysis, TNK + EVT was associated with borderline increased parenchymal hematoma risk (OR 1.54, p = 0.048) but not sICH.


Criticisms

  • Retrospective pooled registry design with unmeasured confounding that PSOW cannot fully address
  • Two treatment groups drawn from different patient populations and largely different stroke centers, increasing residual confounding risk despite weighting
  • Outcomes assessed unblinded in real-world conditions, introducing potential mRS misclassification bias
  • Pre-EVT recanalization data unavailable for most patients, limiting mechanistic interpretation
  • Patients enrolled over 9 years; temporal changes in clinical practice may introduce time-period bias
  • Expected onset-to-thrombolysis time in the EVT-alone group was derived (not directly observed), introducing estimation error
  • mRS at hospital discharge not routinely collected, precluding use for imputing missing 3-month mRS values
  • Brain imaging assessed locally (not centrally), potentially introducing site variability
  • TNK + EVT group had longer onset-to-reperfusion times, which may have partially offset the benefit of tenecteplase
  • Declared reasons for IVT omission in ETIS may be incomplete, and unrecorded contraindications could introduce selection bias

Funding

DMU Neurosciences, APHP, Sorbonne Universite; Investissements d'avenir ANR-10-IAIHU-06; TETRIS registry funded by Boehringer Ingelheim

Based on: ETIS-TETRIS (Neurology, 2026)

Authors: Gaspard Gerschenfeld, Bertrand Lapergue, Pierre Seners, ..., Sonia Alamowitch

Citation: Neurology 2026;106:e214702

Content summarized and formatted by NeuroTrials.ai.