Tenecteplase versus alteplase

A trace: from how the trials measured recovery, up to a Class 1 guideline built on non-inferiority.

Paulina Del Mundo

2026-06-10

For two decades the answer to “which clot-busting drug for an acute ischemic stroke” was settled: alteplase, the standard since 1995. Tenecteplase, a genetically modified cousin given as a single bolus rather than a bolus plus an hour-long infusion, was the challenger. In January 2026 the challenger drew level. The AHA/ASA guideline now carries a Class 1 recommendation to use either drug within 4.5 hours. A drug that was the “reasonable alternative” became a co-equal first-line option, and some stroke neurologists now argue it should be preferred.

This is a trace. It starts at the bottom of the pathway, with how the trials actually measured recovery, and builds back up to that Class 1 sentence, appraising the statistics at each rung. The question it keeps in view is simple: the recommendation says the two drugs are interchangeable and one may be better to give, but what kind of evidence is that claim resting on?

From the datapoint up

01 — Measurement

How recovery was measured

The outcome that matters to a stroke patient is disability, and the trials measure it with the modified Rankin Scale (mRS), a 0-to-6 scale from no symptoms to death, scored at 90 days by an assessor blinded to treatment. That blinding is the quality safeguard in the open-label “PROBE” designs most of these trials used.

The subtlety is in how the scale gets cut. Some trials analyze the whole ordinal distribution, a “shift” across every level of the scale. Others dichotomize at a single point: excellent recovery (mRS 0–1) or the looser functional independence (mRS 0–2). The cut point decides what counts as success, and it changes the statistical power and the answer. One trial in this literature, EXTEND-IA TNK, did not use disability as its primary outcome at all. It used an angiographic reperfusion surrogate, whether the blocked artery reopened, which is a different and earlier question than how the patient is doing three months later. Carry that up: “recovery” is measured several ways, and the comparisons are not all on the same scale.

02 — Model

What the trials tested, and how

Most of the modern evidence is built on non-inferiority, not superiority. That is the single most important design fact in this trace. A non-inferiority trial does not try to show the new drug is better. It tries to show it is not worse by more than a pre-specified margin the investigators choose in advance.

ATTEST-2, the largest precise head-to-head, randomized 1,858 patients (1,777 analyzed) to tenecteplase 0.25 mg/kg or alteplase 0.9 mg/kg within 4.5 hours, and pre-specified a non-inferiority margin of an odds ratio of 0.75 on the ordinal mRS shift, tested hierarchically: non-inferiority first, then superiority only if that passed. Two other large trials at the now-standard 0.25 mg/kg dose, AcT (2022) and TRACE-2 (2023), ran the same non-inferiority comparison with 90-day functional primary outcomes, and together with ATTEST-2 they form the evidence base the 2026 upgrade rests on.

Two trials sit apart and are often miscited. NOR-TEST (2017) was a superiority trial, and it used the higher 0.4 mg/kg dose in a population that turned out to be mostly mild strokes. EXTEND-IA TNK studied large-vessel-occlusion patients headed for thrombectomy, on the reperfusion surrogate. Different designs, different doses, different patients. The “tenecteplase works” headline is really a stack of different questions.

03 — Estimate

The numbers

Start with the precise anchor. In ATTEST-2 the common odds ratio for the mRS shift was 1.07, with a 95% confidence interval of 0.90 to 1.27. Non-inferiority was met decisively (p < 0.0001). Superiority was explicitly not met (p = 0.43). Safety was a wash: symptomatic intracranial hemorrhage in 2% of each arm, and 90-day mortality 8% in each arm. The honest one-line reading is not worse, not better, equally safe.

The trials that sit apart tell the rest. NOR-TEST, at the higher 0.4 mg/kg dose in mild strokes, found nothing: excellent-recovery odds ratio 1.08 (95% CI 0.84 to 1.38, p = 0.52). EXTEND-IA TNK is the one place tenecteplase looked clearly better, but on the reperfusion surrogate, in large-vessel occlusions: the artery reopened in 22% versus 10% (incidence ratio 2.2, 95% CI 1.1 to 4.4, p = 0.03 for superiority). Its 90-day disability benefit was a secondary endpoint with a confidence interval touching the no-effect line (common odds ratio 1.7, 95% CI 1.0 to 2.8), in a narrow population. A companion dose-comparison trial settled a related question: the higher 0.4 mg/kg dose bought no extra reperfusion over 0.25 (19.3% in each arm), which is part of why 0.25 became the standard.

04 — Synthesis

From trials to “the evidence”

Pooled across trials, the picture converges on equivalence. A 2024 meta-analysis of nine trials (3,573 patients) found no significant difference on any functional endpoint: functional independence (mRS 0–2) risk ratio 1.12 (95% CI 0.96 to 1.31), excellent recovery (mRS 0–1) risk ratio 1.04 (0.92 to 1.17), any intracranial hemorrhage risk ratio 1.14 (0.77 to 1.68), 90-day mortality risk ratio 0.97 (0.72 to 1.29). Every interval crosses 1. Tenecteplase is neither clearly better nor clearly worse.

Two cautions belong here. A dose-stratified safety analysis found symptomatic hemorrhage comparable to alteplase at 0.25 mg/kg (risk ratio 0.77, 95% CI 0.53 to 1.14) but trending higher at 0.4 mg/kg (risk ratio 2.31), the interval wide and crossing 1, which again favors the lower dose. And a 2025 meta-analysis restricted to large-vessel occlusions found excellent recovery right at the edge of significance (risk ratio 1.18, 95% CI 1.00 to 1.40, p = 0.05) in a small, wide-CI sample. The synthesis verdict is as good as, about as safe, with a faint lean on the strictest endpoint that never quite reaches significance.

05 — Decision rule

The dose and the clock

The rule a clinician acts on is compact: tenecteplase 0.25 mg/kg as a single intravenous bolus (maximum 25 mg), within 4.5 hours of onset. The 0.4 mg/kg dose was set aside, no added benefit and a safety signal.

The decisive practical fact at this rung is not an effect size. Alteplase is a bolus followed by a 60-minute infusion that has to be maintained, including during transfer. Tenecteplase is one push. For a patient being moved from a community hospital to a thrombectomy center, a single bolus is faster to give, simpler to manage, and easier to hand off. That logistical advantage is real, and it is much of why clinicians want to make the switch. It is also a different kind of reason than a better outcome.

06 — Recommendation

“Either tenecteplase or alteplase,” Class 1

In January 2026 the AHA/ASA acute ischemic stroke guideline (Prabhakaran et al., Stroke 2026) issued a Class 1 recommendation to use either tenecteplase 0.25 mg/kg or alteplase 0.9 mg/kg within 4.5 hours to improve functional outcomes. That is the guideline’s strongest tier, and it puts the two drugs on equal footing.

It is also a clear upgrade. The 2019 update had placed tenecteplase only at Class IIb, the “may be reasonable” tier, and largely in patients already headed for thrombectomy. In one trial report the investigators went further than any guideline, writing that tenecteplase “should be preferred,” on the strength of its single-bolus delivery. So across seven years the recommendation moved from alternative, to co-equal first-line, to preferred in some expert hands.

The verdict

Stand at the top and look back down.

The estimate at the center of the case is a non-inferiority result. ATTEST-2, the largest precise head-to-head, met its pre-specified margin and explicitly did not show superiority (p = 0.43); the pooled data agree that every functional interval crosses 1. This is genuinely reassuring evidence. It is reassurance of a specific kind: a non-inferiority result bounds how much worse a drug could plausibly be. It cannot, on its own, establish that the drug is better.

So the step from what was shown to what was recommended is where the language runs slightly ahead of the design. The Class 1 endorsement, and the stronger “should be preferred,” rest on two things: equivalent efficacy within a margin the investigators chose, and a real logistical advantage in giving a single bolus. That is a legitimate and even strong basis for recommending a drug. Equivalent outcomes plus easier, faster administration is a true clinical advantage. But it is a different claim from “tenecteplase produces better outcomes,” and the all-comers data do not support that stronger claim. The one place tenecteplase looked superior was a reperfusion surrogate, in large-vessel occlusions, with the downstream disability benefit a borderline secondary endpoint. It should not be read as a general efficacy edge.

None of this argues against tenecteplase. Equal effectiveness with a single bolus is a good reason to use it, and may be a good reason to prefer it. It argues for naming the claim correctly: a Class 1 recommendation resting on non-inferiority and logistics, not on demonstrated superiority.

The non-inferiority margin is the quiet load-bearing choice. An odds ratio of 0.75 is the amount of “worse” the trial agreed in advance to tolerate and still call a success. Move that margin, and the same data could read as a closer or a more permissive call. A reader who takes “Class 1, either agent” at face value is trusting that margin without ever seeing it.

The contrast that bounds it

The cleanest check on the headline is to watch it change with dose and population. NOR-TEST ran a superiority test at the higher 0.4 mg/kg dose in mostly mild strokes and found nothing, with a hint of more bleeding at that dose. EXTEND-IA TNK ran tenecteplase at 0.25 mg/kg in large-vessel occlusions and found a clear benefit on reopening the artery, a surrogate. The recommended-dose, all-comers truth sits between them: non-inferior to alteplase, about as safe, not proven better.

Two doses, several populations, three different primary outcomes, one compressed guideline sentence. That compression is exactly what a trace exists to unpack: “either agent, Class 1” is well supported as a statement about non-inferiority and logistics, and it quietly carries more than that when it is read as “the drugs are the same, and the newer one is better to give.”

This trace is methodological commentary, not clinical advice. It is an appraisal of what the statistics behind a recommendation can and cannot support, written for people who design studies, defend methods, and read the literature critically. It is not guidance for treating a patient, and it is not a substitute for the guidelines themselves or for clinical judgment.

It is also part of a fast-moving field: the trial figures here are historical and stable, but guideline language continues to evolve, and the exact level of evidence behind the 2026 Class 1 recommendation should be read in the source.

If you have a recommendation, a guideline-development question, or a manuscript whose statistical basis you want traced this way, that is the kind of work I take on. Book a discovery call →

This trace is part of Acute ischemic stroke (I63), on the From Data to Bedside pathway.

Sources

  • Muir KW, et al. (ATTEST-2 Investigators). Tenecteplase versus alteplase for acute ischaemic stroke (ATTEST-2): a randomised, parallel-group, open-label, blinded-endpoint, non-inferiority trial. Lancet Neurol. 2024. https://pubmed.ncbi.nlm.nih.gov/39424558/
  • Menon BK, et al. Intravenous tenecteplase compared with alteplase for acute ischaemic stroke in Canada (AcT): a pragmatic, multicentre, open-label, registry-linked, randomised, controlled, non-inferiority trial. Lancet. 2022. https://www.thelancet.com/article/S0140-6736(22)01054-6/fulltext
  • Wang Y, et al. Tenecteplase versus alteplase in acute ischaemic stroke within 4.5 h (TRACE-2): a phase 3, multicentre, open-label, randomised controlled, non-inferiority trial. Lancet. 2023.
  • Logallo N, et al. Tenecteplase versus alteplase for management of acute ischaemic stroke (NOR-TEST): a phase 3, randomised, open-label, blinded endpoint trial. Lancet Neurol. 2017. https://pubmed.ncbi.nlm.nih.gov/28780236/
  • Campbell BCV, et al. Tenecteplase versus Alteplase before Thrombectomy for Ischemic Stroke (EXTEND-IA TNK). N Engl J Med. 2018. doi:10.1056/NEJMoa1716405.
  • Campbell BCV, et al. Effect of Intravenous Tenecteplase Dose on Cerebral Reperfusion Before Thrombectomy (EXTEND-IA TNK Part 2). JAMA. 2020. https://jamanetwork.com/journals/jama/fullarticle/2761799
  • Dose-stratified safety meta-analysis of tenecteplase for acute ischemic stroke. Stroke. 2023. doi:10.1161/STROKEAHA.122.042335.
  • Singh S, et al. Tenecteplase versus alteplase for acute ischemic stroke: a systematic review and meta-analysis. 2024. https://pubmed.ncbi.nlm.nih.gov/38694719/
  • Large-vessel-occlusion meta-analysis of tenecteplase versus alteplase. Front Neurol. 2025. doi:10.3389/fneur.2025.1487711.
  • Prabhakaran S, Gonzalez NR, Zachrison KS, et al. 2026 Guideline for the Early Management of Patients With Acute Ischemic Stroke. Stroke. Published online January 26, 2026.
  • Powers WJ, et al. Guidelines for the Early Management of Patients With Acute Ischemic Stroke: 2019 Update. Stroke. 2019;50:e344–e418. doi:10.1161/STR.0000000000000211.