Quick answer
Thrombectomy transformed large-vessel stroke, but most ischaemic strokes are not large-vessel occlusions — they involve medium vessels, intracranial stenosis or other mechanisms, and thrombectomy does not apply. For these patients intravenous thrombolysis within 4.5 hours is standard, and beyond that window there was little evidence. OPTION, led by Hao Junwei at Xuanwu Hospital across 48 Chinese centres, tested tenecteplase at 4.5 to 24 hours in patients with salvageable tissue on perfusion imaging. Excellent 90-day outcome rose from 34.2% to 43.6%. It was presented at the International Stroke Conference and published in JAMA on 5 February 2026 [1][3][5].
Key takeaways
- Design: multicentre, prospective, randomised, open-label, blinded-endpoint trial at 48 Chinese centres (NCT05752916) [5].
- Population: 570 randomised, 566 analysed; non-large-vessel occlusion, 4.5–24 hours, perfusion mismatch required [3][5].
- Primary outcome: 90-day mRS 0–1 in 43.6% with tenecteplase versus 34.2% with standard medical therapy [4][6].
- Safety: symptomatic intracranial haemorrhage at 36 hours 2.8% versus 0%; 90-day mortality 5.0% versus 3.2% [6].
- Authors: Ma Gaoting, Mo Ran, Zuo Yingting and Ma Qingfeng co-first; Hao Junwei corresponding [2].
Why the imaging filter matters more than the clock
The trial did not simply extend a time window; it replaced time with tissue. Eligible patients needed a perfusion mismatch — core under 50 mL, hypoperfusion-to-core ratio at least 1.2, and a mismatch volume of at least 10 mL. In other words, the brain had to still contain tissue that was starved but not yet dead [5].
That selection is what makes the result plausible: in a patient whose core has already completed, reopening a small vessel adds bleeding risk without salvageable benefit. It also means the finding is not an argument for treating everyone who presents late — it is an argument for imaging them [3][5].
Weighing a 9-point gain against a bleeding signal
A 9.4 percentage-point absolute increase in excellent outcome is clinically substantial — comparable to what thrombectomy delivers in its population. But symptomatic intracranial haemorrhage occurred in 2.8% versus 0% of controls, and 90-day mortality was numerically higher at 5.0% versus 3.2%, although the trial was not powered to adjudicate mortality [4][6].
Clinicians knew which patients received thrombolysis, and outcome assessors were blinded; that is an accepted design but not a perfect one. The 2.8% haemorrhage rate is within the range seen in early-window thrombolysis, yet against a zero-event control arm it deserves attention, and the mortality direction should be tracked in longer follow-up and meta-analysis. The trial enrolled patients with mild-to-moderate deficits (NIHSS 6–25, or 4–5 with disabling deficits), so results do not extend to very mild or very severe presentations [5][6].
The condition in context: ischaemic stroke
Stroke is the leading cause of adult death and disability in China, and late presentation is common.
| Region | What the data shows | Source |
|---|---|---|
| Case mix | Most ischaemic strokes are not large-vessel occlusions and are ineligible for thrombectomy [1][2]. | Xuanwu Hospital [1][2] |
| Prior evidence | Late-window thrombolysis evidence existed for large-vessel occlusion but not for this group [5]. | CMT [5] |
| Practical implication | CT perfusion capability determines which hospitals can apply the result [5]. | CMT [5] |
Because many patients arrive beyond 4.5 hours, an effective late-window option addresses the majority of real-world presentations [2][3].
Why it matters
| Potential | What the report suggests | Source |
|---|---|---|
| New option | Thrombolysis extended to 24 hours with imaging selection [4][6]. | [4][6] |
| Effect size | 43.6% versus 34.2% excellent outcome [4]. | [4] |
| Caution | Haemorrhage 2.8% versus 0%; open-label [6]. | [6] |
Availability and limits
- This is an editorial summary of the publication, not a treatment recommendation [1][5].
- Requires CT perfusion and a centre able to act on it [5].
- Applies to non-large-vessel occlusion with mismatch, not to all late presentations [3].
- Bleeding risk requires patient-level judgement [6].
- Longer-term mortality and functional data are still maturing [4].
What to ask before booking
| Question | Why it matters |
|---|---|
| Can this hospital do CT perfusion out of hours? | The eligibility criterion depends on it [5]. |
| Do I have a large-vessel occlusion? | That changes the treatment to thrombectomy [2]. |
| Which department handles foreign patients, and in what language? | Xuanwu is a working hospital we do not book; confirm international-department coordination and language support. |
| What is my individual bleeding risk? | The benefit-haemorrhage trade-off is patient-specific [6]. |
Questions patients ask
What did OPTION test?
Tenecteplase versus standard medical therapy in acute non-large-vessel occlusion stroke presenting 4.5–24 hours after onset, with salvageable tissue confirmed on CT perfusion [3][5].
What was the result?
90-day excellent outcome (mRS 0–1) was 43.6% with tenecteplase versus 34.2% with standard therapy [4].
Was it safe?
Symptomatic intracranial haemorrhage was 2.8% versus 0%, and 90-day mortality 5.0% versus 3.2%; the trade-off needs individual assessment [6].
Where was it published?
JAMA, online 5 February 2026, and presented at the International Stroke Conference on 6 February 2026 [1][2].
Is this relevant to international patients?
It is useful context for stroke care in China. Xuanwu is a working hospital we do not book — contact its international department directly [1].