China's Tenecteplase Trial Challenges the 4.5-Hour Thrombolysis Wall — and Exposes a U.S. Regulatory Gap

Picture a stroke unit in China at 2 a.m. A patient arrives with confirmed proximal middle cerebral artery occlusion — the imaging shows a dense clot, the clock shows 18 hours since symptom onset. Under the conventional Western protocol, that patient goes straight to the angiography suite for endovascular treatment. Under the trial protocol published in JAMA, that same patient first receives an intravenous bolus of tenecteplaseView drug details. The next 90 days will determine whether that decision was correct — or catastrophic.

The trial's finding: it was correct. The phase 3 randomized superiority trial, conducted across sites in China, evaluated tenecteplaseView drug details administered prior to endovascular treatment versus endovascular treatment alone in patients presenting between 4.5 and 24 hours after ischemic stroke onset due to proximal MCA occlusion. The results landed on the right side of the superiority threshold. But the more important story involves what this trial reveals about how rigorous CNS trial design actually operates — and what the U.S. regulatory ecosystem has been slow to absorb.

The Window No One Trusted to Open

For years, the clinical assumption governing thrombolysis in stroke was essentially this: the earlier you treat, the safer the treatment; the later you treat, the more you risk hemorrhagic transformation. That assumption was not wrong, but it was incomplete in ways that trial design has consistently failed to challenge.

The DAWN trial, conducted under an adaptive design and cited in American Heart Association Journals as a landmark for adaptive methodology in stroke, extended thrombectomy eligibility out to 24 hours using perfusion imaging to identify patients with salvageable brain tissue. DAWN did not include a thrombolysis arm. The underlying logic — that imaging selection could identify who benefits beyond conventional time windows — was validated for mechanical intervention but left thrombolysis largely frozen at the 4.5-hour wall. This Chinese phase 3 trial is the first rigorous superiority-designed RCT to directly challenge that wall for tenecteplaseView drug details in the context of EVT.

The protocol design reflects lessons the field learned painfully. Proximal MCA occlusion was not incidental — it was the inclusion criterion that made the science tractable. Patients with proximal MCA occlusion have a well-characterized natural history, a defined imaging phenotype on CTA and MRI perfusion, and a biological rationale for why partial thrombolysis before mechanical retrieval might improve first-pass recanalization. The endpoints were pre-registered, the safety monitoring included hemorrhagic transformation as a primary adverse event, and the primary efficacy outcome was functional independence at 90 days — the mRS endpoint that stroke trials have standardized precisely because regulators on both sides of the Pacific accept it.

The FDA Approval Gap

GenentechView company profile's March 3, 2025 FDA approval of TNKaseView drug details® (tenecteplaseView drug details) for acute ischemic stroke (search)Search disease in adults cleared a single five-second intravenous bolus replacing alteplaseView drug details's 60-minute infusion. That approval covered patients eligible for IV thrombolysis under established criteria. What it did not cover was the late-window, pre-EVT combination strategy this trial tested. The gap between what the FDA approved and what this phase 3 trial evaluated is precisely where the next label expansion debate will be fought.

The ATTEST trial, which compared tenecteplaseView drug details against alteplaseView drug details in acute ischemic stroke (search)Search disease, reported symptomatic intracerebral hemorrhage by the SITS-MOST definition at 2% in the tenecteplase arm — but that was a population treated promptly, not one stratified by a late-presenting imaging profile. Extending the window to 24 hours while adding the constraint of proximal MCA occlusion is a different scientific question entirely, and this trial treated it as one.

What the Protocol Got Right

A meta-analysis published in April 2026 covering 21 studies — 8 randomized controlled trials and 13 observational studies — with 10,538 total patients found broadly that thrombolysis before thrombectomy in large vessel ischemic stroke carries a manageable safety profile, but the heterogeneity across those studies has made regulatory synthesis difficult. Different definitions of symptomatic hemorrhage, different time windows, different EVT protocols, and different imaging eligibility criteria mean that pooling those studies tells you something about the direction of effect but almost nothing you can take into a Type B pre-submission meeting and use to anchor a label expansion argument.

The EXTEND-IA TNK trials — which produced a pooled analysis examining thrombus characteristics associated with early reperfusion — were groundbreaking for establishing the mechanistic basis of tenecteplaseView drug details-before-thrombectomy. But EXTEND-IA TNK was designed in the early window, and its sample sizes, while statistically powered, were not constructed to support a superiority claim in a late-presenting population. The Chinese trial was. That is not a marginal distinction — it is the structural difference between a hypothesis-generating dataset and a submission-ready evidentiary package.

The AHA/ASA's Target: Stroke Phase III initiative sets a benchmark that 85% or more of IV thrombolytic patients should achieve door-to-needle times within 60 minutes. In the context of a trial that also requires EVT team activation, imaging review for MCA occlusion confirmation, and tenecteplaseView drug details administration before arterial access, the operational choreography is demanding. Sites that met those timing requirements while maintaining protocol fidelity in a multi-center Chinese study demonstrated something operationally important: the bridging therapy workflow is executable at scale, not just in academic centers with dedicated stroke programs.

The Regulatory Blind Spot

Sponsors seeking a U.S. label expansion for tenecteplaseView drug details into the late-window pre-EVT setting will face an immediate structural problem. The FDA approved TNKaseView drug details on March 3, 2025 for a population defined by conventional thrombolysis eligibility — not by imaging-selected late-window proximal MCA occlusion. Using this Chinese phase 3 trial as the primary evidentiary basis for an sNDA or supplemental BLA will require FDA acceptance of a dataset generated entirely outside the United States, under NMPA oversight, with a patient population whose imaging protocols, site infrastructure, and EVT technical capabilities differ meaningfully from U.S. practice.

The NMPA has accelerated approvals in ischemic stroke before — the approval of edaravone and dexborneol sublingual tablets under an expedited pathway reflects the agency's willingness to move on CNS indications when the evidence package is organized correctly. But the FDA's reciprocal recognition of trials conducted under NMPA oversight remains inconsistent. ICH E17, the multi-regional clinical trial guideline, provides the framework for arguing that a China-only trial generates data generalizable to a U.S. population — but the argument requires a sponsor willing to make it, with an FDA division willing to engage on it, in a disease area where the evidentiary bar has historically been set by U.S. and European trial data.

Consider what happens when a sponsor brings this trial to a Type B meeting request for a late-window tenecteplaseView drug details pre-EVT sNDA. The FDA reviewer's first question will be about the patient population's representativeness. The second will be about the imaging modality standardization across Chinese sites. The third — and most consequential — will be about whether the EVT techniques used in Chinese centers are sufficiently comparable to U.S. practice to assume the combination effect is generalizable. None of these questions have obvious answers, and none of them have been formally addressed in FDA guidance on stroke trial design in the late-window setting.

That absence of guidance is the real operational problem. Sponsors looking to move tenecteplaseView drug details into this indication for U.S. patients do not have a regulatory roadmap. The FDA's 2019 guidance on analytic approaches to support early feasibility studies and the broader adaptive design guidance issued in 2019 address methodology — but neither speaks to the specific evidentiary requirements for a thrombolytic label expansion in an imaging-defined stroke subpopulation.

The Path Forward

For sponsors and their regulatory teams, the operational move is to request a Type B pre-submission meeting before investing in a U.S. bridging study — and to come to that meeting with a specific proposal under ICH E17 for how the Chinese phase 3 data satisfies the substantial evidence standard. The alternative — conducting a parallel U.S. trial in this narrow population — is not operationally trivial. Proximal MCA occlusion in the 4.5-to-24-hour window represents a fraction of stroke presentations, and site activation for a superiority-powered RCT in this subgroup would require a network of comprehensive stroke centers with aligned EVT protocols, imaging infrastructure, and enrollment capacity that does not currently exist as a coordinated research network in the United States.

Back in that Chinese stroke unit at 2 a.m., the trial coordinator who randomized that late-presenting patient was operating inside a protocol that answered a question the FDA's current label has not yet had to ask. The 90-day mRS data from that patient — and the thousands like her enrolled across this trial — now sit in a JAMA publication that every stroke neurologist in the United States will read. Whether that data eventually reshapes the TNKaseView drug details label depends entirely on whether a sponsor decides the regulatory path is worth walking, and whether the FDA decides the science is worth engaging on its own terms. The trial made the scientific case. The regulatory case is still waiting to be built.