← Back

EXTEND-IA TNK

Tenecteplase versus Alteplase before Thrombectomy for Ischemic Stroke

Share Share Copied! Compare

Year of Publication: 2018

Authors: Bruce C.V. Campbell, Peter J. Mitchell, Leonid Churilov, ..., Vincent Thijs

Journal: New England Journal of Medicine

Citation: N Engl J Med 2018;378:1573–1582. DOI:10.1056/NEJMoa1716405

Link: https://www.nejm.org/doi/full/10.1056/NEJMoa1716405

PDF: https://www.nejm.org/doi/pdf/10.1056/NEJMoa1716405

Clinical Question

Is intravenous tenecteplase superior or noninferior to alteplase before endovascular thrombectomy in achieving reperfusion and improving outcomes in acute ischemic stroke?

Study Overview

Objective

To compare the efficacy and safety of tenecteplase versus alteplase in achieving early reperfusion prior to endovascular thrombectomy in acute ischemic stroke.

Study Summary

Tenecteplase was associated with higher early reperfusion rates and improved functional outcomes compared to alteplase.

Intervention

Randomized administration of IV tenecteplase (0.25 mg/kg, max 25 mg) as a single bolus vs. IV alteplase (0.9 mg/kg, max 90 mg, 10% bolus + 1-hour infusion) within 4.5 hours of stroke onset; arterial puncture for endovascular thrombectomy could commence within 6 hours of onset.

Patients per Arm

Tenecteplase: 101, Alteplase: 101

Bottom Line

Tenecteplase led to higher rates of early reperfusion and improved functional outcomes compared to alteplase in patients undergoing thrombectomy within 4.5 hours of symptom onset.

        Major Points
        EXTEND-IA TNK was the FIRST randomized trial to demonstrate tenecteplase superiority over alteplase before thrombectomy — launching the global shift toward tenecteplase as the preferred thrombolytic for acute stroke.
202 patients with LVO (ICA, M1, M2, or basilar) eligible for IV thrombolysis within 4.5 hours of onset and arterial puncture for thrombectomy within 6 hours of onset, randomized at 12 Australian and 1 New Zealand centers. Primary outcome: substantial early reperfusion (≥50% territory or no thrombus) on initial angiography.
Tenecteplase DOUBLED early reperfusion: 22% vs 10% (incidence difference 12 percentage points, 95% CI 2–21, P=0.002 for noninferiority; adjusted incidence ratio 2.2, 95% CI 1.1–4.4, P=0.03 for superiority; adjusted OR 2.6, 95% CI 1.1–5.9, P=0.02). This means more patients arrived at the cath lab with partially or completely dissolved clots — simplifying the thrombectomy procedure.
Better 90-day outcomes: median mRS 2 vs 3 (common OR 1.7, 95% CI 1.0–2.8, P=0.04). Functional independence (mRS 0–2) trended higher: 64% vs 51% (adjusted OR 1.8, 95% CI 1.0–3.4, P=0.06). Mortality trended lower: 10% vs 18% (adjusted risk ratio 0.5, 95% CI 0.3–1.0, P=0.049; adjusted OR 0.4, 95% CI 0.2–1.1, P=0.08).
Safety was equivalent: sICH was 1/101 (1%) in BOTH groups (risk ratio 1.0, 95% CI 0.1–15.9, P=0.99) — identical hemorrhage rates despite higher recanalization, disproving concerns that tenecteplase might increase bleeding risk.
Practical advantage: tenecteplase is given as a single IV BOLUS over 5–10 seconds (weight-based, 0.25 mg/kg, max 25 mg) — no 1-hour infusion like alteplase. This is transformative for 'drip-and-ship' models where patients need immediate transfer after thrombolysis.
Tenecteplase 0.25 mg/kg dose was selected based on prior stroke data (Parsons 2012) showing better outcomes than 0.1 mg/kg. This dose became the standard in subsequent trials (AcT, TASTE, NOR-TEST 2).
Pharmacology advantage: tenecteplase has ~20-minute half-life (vs alteplase's ~4 min), 14× higher fibrin specificity, and 80× greater resistance to PAI-1 inhibition — making it a superior thrombolytic from a pharmacokinetic standpoint.
Led directly to EXTEND-IA TNK Part 2 (0.25 vs 0.40 mg/kg dose comparison), AcT (Canadian phase 3, positive), and ultimately guideline endorsement of tenecteplase as an acceptable alternative to alteplase (2019 AHA/ASA update).
Australia/NZ became the first region to adopt tenecteplase as first-line for acute stroke — demonstrating that a single well-designed phase 2 trial can shift global practice when the pharmacologic rationale is strong.

      

Design

Study Type: Multicenter, randomized, open-label, blinded-endpoint trial (PROBE design)

Randomization: 1

Blinding: Blinded outcome assessment

Enrollment Period: March 2015 – October 2017

Follow-up Duration: 90 days

Centers: 13

Countries: Australia, New Zealand

Sample Size: 202

Analysis: Intention-to-treat and per-protocol analyses (204 enrolled, 2 excluded before treatment — 1 withdrawal of consent, 1 physician withdrawal for eligibility error; no patients excluded from the per-protocol analysis of the primary outcome, so only one set of analyses is presented); adjusted for baseline NIHSS, age, and site of vessel occlusion

Inclusion Criteria

Ischemic stroke with large-vessel occlusion (ICA, MCA M1, MCA M2, or basilar artery)
Eligible for IV thrombolysis within 4.5 hours of symptom onset
Arterial puncture for endovascular thrombectomy could commence within 6 hours of onset
Originally required CT-perfusion mismatch for anterior-circulation strokes (mismatch ratio >1.2, absolute volume difference >10 ml, ischemic-core volume <70 ml) — mismatch criteria removed on October 12, 2016 after approximately 80 patients enrolled

Exclusion Criteria

Pre-stroke modified Rankin scale >3
Contraindications to thrombolysis or thrombectomy

Baseline Characteristics

Characteristic	Comorbidities	Qualifying Event

Arms

Field	Tenecteplase	Control
Intervention	Tenecteplase 0.25 mg/kg IV bolus (max 25 mg) before thrombectomy	Alteplase 0.9 mg/kg IV (max 90 mg, with 10% bolus + 1-hour infusion)
Duration	Single dose, followed by thrombectomy	Standard infusion prior to thrombectomy

Outcomes

Outcome	Type	Control	Intervention	HR / OR / RR	P-value
Substantial reperfusion (>50% of involved territory or absence of retrievable thrombus) at initial angiographic assessment	Primary	10/101 (10%)	22/101 (22%)	9	P=0.002 (noninferiority); P=0.03 (adjusted incidence ratio, superiority); P=0.02 (adjusted OR, superiority)
Median mRS at 90 days (ordinal analysis, common OR 1.7, 95% CI 1.0–2.8)	Secondary	3 (IQR 1–4)	2 (IQR 0–3)	Common OR 1.7 (1.0–2.8)	0.04
Functional independence — mRS 0–2 or no change from baseline (adjusted incidence ratio 1.2, 1.0–1.5; adjusted OR 1.8, 1.0–3.4)	Secondary	52/101 (51%)	65/101 (64%)	Adjusted OR 1.8 (1.0–3.4)	0.06
Excellent outcome — mRS 0–1 or no change from baseline (adjusted incidence ratio 1.2, 0.9–1.6; adjusted OR 1.4, 0.8–2.6)	Secondary	43/101 (43%)	52/101 (51%)	Adjusted OR 1.4 (0.8–2.6)	0.23
Early neurologic improvement at 72h (≥8-point NIHSS drop or NIHSS 0–1)	Secondary	69/101 (68%)	72/101 (71%)	Adjusted OR 1.1 (0.6–2.1)	0.70
Symptomatic ICH	Adverse	1/101 (1%)	1/101 (1%)	Risk ratio 1.0 (0.1–15.9); OR 1.0 (0.1–16.2)	0.99
Death	Adverse	18/101 (18%)	10/101 (10%)	Adjusted risk ratio 0.5 (0.3–1.0); adjusted OR 0.4 (0.2–1.1)	P=0.049 (risk ratio); P=0.08 (OR)
Parenchymal hematoma	Adverse	5/101 (5%)	6/101 (6%)	Risk ratio 1.2 (0.4–3.8); OR 1.2 (0.4–4.1)	0.76

Criticisms

Open-label design with blinded endpoint assessment — awareness of thrombolytic type could influence clinical decisions (e.g., timing of thrombectomy).
Originally powered for noninferiority, with sequential testing of superiority — the superiority findings for the primary reperfusion outcome should be interpreted with appropriate caution.
Small sample size (n=202) limits statistical power for secondary clinical outcomes (mRS 0–2 difference did not reach significance, P=0.06).
CT-perfusion mismatch selection criteria were removed mid-trial (October 12, 2016) after approximately 80 patients enrolled, changing the eligible population and raising concerns about selection consistency.
Enrolled at 12 Australian centers plus 1 New Zealand center — practice patterns (rapid workflow, high thrombectomy rates) may not generalize to all healthcare systems.
Thrombectomy was not performed in patients who achieved reperfusion on the initial angiogram (except 1 tenecteplase patient with residual thrombus), so the trial cannot isolate the benefit of tenecteplase in patients who ultimately did not undergo thrombectomy.
The 0.25 mg/kg dose was selected based on prior stroke data (Parsons 2012) showing better outcomes than 0.1 mg/kg; EXTEND-IA TNK Part 2 later showed 0.40 mg/kg offered no additional benefit, but neither dose was compared to no thrombolysis.
Short enrollment period and relatively small number of centers may introduce site-selection bias toward high-volume stroke centers.
The primary endpoint (reperfusion on initial angiogram) is a surrogate marker — clinical outcomes (mRS) showed favorable trends but did not all reach significance.

Funding

National Health and Medical Research Council of Australia (grants 1043242, 1035688, 1113352, 1111972); Royal Australasian College of Physicians; Royal Melbourne Hospital Foundation; National Heart Foundation of Australia; Stroke Foundation of Australia; infrastructure funding from the state government of Victoria; unrestricted grant from Medtronic.

Based on: EXTEND-IA TNK (New England Journal of Medicine, 2018)