Armghan Ans: Striking Evidence Reversals of The Last Decade for Stroke

Armghan Ans, Assistant Professor and Director of Stroke at UPMC Washington, Founder of MDAdopt, shared on LinkedIn:

”Stroke intervention just went through one of the more striking evidence reversals of the last decade.

In a fifteen-month span, the field’s verdict on medium vessel occlusion (MeVO) thrombectomy flipped from ‘does not work’ to ‘works in the right patients.’

Three trials in early 2025 said no.

ORIENTAL-MeVO in 2026 said yes – but only under conditions the 2025 trials largely did not test.

What changed?

Not the science.

The selection.

Full breakdown of the 2025-to-2026 arc, the LVO parallel from a decade ago, and what it means for hospitals, imaging, and workflow in the newsletter:

What 2025 Said, What 2026 Changed, and What It Means for the Field

Stroke intervention has just gone through one of the more striking evidence reversals of the last decade.

In a fifteen-month span, the field’s verdict on medium vessel occlusion thrombectomy flipped from ‘does not work’ to ‘works in the right patients.’

Three trials in early 2025 said no. One trial in 2026, plus a device-specific trial alongside it, said yes.

But only under conditions that the 2025 trials largely did not test.

The science did not change. The selection did.

The 2025 Verdict

Three randomized trials reported within days of each other at the International Stroke Conference in February 2025, and all three pointed the same direction.

ESCAPE-MeVO, the Canadian-led international trial of 530 patients across five countries, found that endovascular thrombectomy did not improve outcomes over best medical therapy and was associated with higher mortality and symptomatic hemorrhage.

DISTAL, the European trial of 543 patients across eleven countries, found a fully neutral result with no benefit on the modified Rankin scale shift.

DISCOUNT, the French multicenter trial across twenty-two hospitals, was halted early on interim analysis when patients in the thrombectomy arm showed significantly worse functional outcomes than those receiving medical management alone.

By mid-2025, the field had three independent investigator groups, three populations, three trial designs, and one direction of effect.

The verdict was not that MeVO thrombectomy was unproven. It was that the available evidence pointed against it.

The 2026 Reversal

ORIENTAL-MeVO, published in NEJM in May 2026, enrolled 563 patients across forty-eight Chinese centers.

The senior investigators included Dr. Raul Nogueira at UPMC, under whom I trained during my vascular neurology fellowship.

The trial found that thrombectomy improved functional independence at 90 days, with a number needed to treat of approximately eight.

This is a clinically meaningful effect, in the range of accepted reperfusion interventions in acute stroke care.

But ORIENTAL-MeVO did not test the same patients the 2025 trials did.

It enrolled moderate-to-severe strokes by design, with stricter operator and center requirements, in M2 occlusions more than M3.

The benefit was concentrated in patients with NIHSS at or above 8, within 8 hours of onset, treated at high-volume centers by experienced operators.

In milder strokes, in M3 occlusions, and beyond 8 hours, the benefit attenuated or disappeared.

The imaging eligibility deserves attention. ORIENTAL-MeVO accepted two qualifying pathways: less than 50% ischemic involvement of the at-risk territory on non-contrast CT or diffusion MRI, or a penumbra-to-core mismatch ratio greater than 1.4 with at least 10 mL of mismatch volume on perfusion imaging.

In practice, roughly half the enrolled patients had baseline perfusion data.

The trial validated perfusion as a qualifying pathway without requiring it universally.

This pragmatic dual standard is itself a statement about where the field is today.

Alongside ORIENTAL-MeVO, the DISTALS trial reported at ISC 2026 using a purpose-built device, the Tigertriever 13, the first stent retriever designed specifically for medium vessel anatomy.

In 118 patients across the United States and Europe, DISTALS showed reperfusion rates above 86% with no symptomatic intracranial hemorrhage in the per-protocol arm.

Functional outcomes at 90 days are still pending, but the procedural safety signal is the cleanest yet seen in this space.

ORIENTAL-MeVO and DISTALS did not show that the 2025 trials were wrong.

They showed that the field had been asking too broad a question.

What Changed

The 2025 and 2026 trials are not contradictory. They are consistent within their own data.

MeVO thrombectomy does not help mild strokes.

It helps moderate-to-severe strokes in selected patients with the right anatomy, the right time window, the right operator, and the right center.

The pattern is familiar. The field traveled a similar arc with large vessel occlusion thrombectomy a decade ago, in two stages.

The 2013 trials were negative. They had enrolled patients largely on the basis of clinical severity alone, often without imaging confirmation of large vessel occlusion.

The 2015 trials – MR CLEAN, ESCAPE, EXTEND-IA, SWIFT PRIME, REVASCAT – were positive.

The intervention had not fundamentally changed in those two years.

CT angiography to confirm large vessel occlusion became a prerequisite for enrollment, and improved stent retrievers replaced first-generation devices.

Three years later, the DAWN and DEFUSE-3 trials extended the time window from 6 to 24 hours by using advanced imaging to identify salvageable tissue as the selection criterion. Imaging had become not just the entry point but the gating mechanism for who got treated, and when.

MeVO thrombectomy did not become effective in 2026. The field became more careful about who we offer it to.

Perfusion as the New Gate?

The shift from ‘does the intervention work’ to ‘in whom does it work’ is not a clinical refinement.

It is an operational transformation. And the operational question now sits squarely on imaging.

When the question was binary – thrombectomy or no thrombectomy for large vessel occlusion – hospitals could build relatively simple selection pathways.

A clinical severity threshold, imaging confirmation of large vessel occlusion, transfer or treat.

The decision sat with one or two clinicians and could be made in minutes.

The MeVO selection question is fundamentally different.

The clinician evaluating a patient with an M2 occlusion now has to weigh stroke severity, vessel segment, time from onset, operator experience, center volume, and imaging mismatch.

The same patient, on the same scanner, can be a thrombectomy candidate or not, depending on how the selection question is constructed.

Advanced imaging is the conceptual and technological substrate for this new selection question, with perfusion as its most prominent modality.

CT perfusion and diffusion MRI were already central to extended-window large vessel occlusion decisions through the DEFUSE-3 and DAWN trials.

ORIENTAL-MeVO extended that logic into medium vessel territory by accepting perfusion mismatch as one qualifying pathway.

The 2026 AHA/ASA stroke guidelines reflect the same direction.

Across the field, advanced imaging is the toolkit that defines who is salvageable, when, and where the line between intervention and observation sits.

The Adoption Gap

The asymmetry is significant.

Academic comprehensive stroke centers generally have twenty-four-hour advanced imaging capability available, with software that processes perfusion or diffusion data into clinically actionable mismatch maps in real time.

Many community hospitals, primary stroke centers, and telestroke spokes do not. Even those with the imaging hardware have not always implemented the workflows that make it actionable.

The MeVO evidence base assumes a level of advanced imaging access and operational maturity that a significant fraction of the stroke care infrastructure does not yet implement.

ORIENTAL-MeVO‘s dual qualifying pathway — perfusion mismatch or non-contrast CT-based ischemic involvement — is partly a recognition of this reality.

The trial treated both as legitimate paths to selection, which has real implications for implementation and adoption across centers with different imaging capabilities.

For imaging AI, this is the adoption gap, and it is more specific than the general detection problem the field has been solving for the last decade.

Detection accuracy is no longer the frontier.

What matters now is whether the platform supports a more nuanced selection question at the bedside, in real time, across a wider range of vessel territories.

Whether the perfusion or diffusion output integrates with the radiology and neurology workflow.

Whether the mismatch map reaches the clinician at the point of decision rather than as an after-the-fact PDF in a folder.

Whether the imaging package speaks the language of the new selection criteria — NIHSS threshold, vessel segment, time window, mismatch ratio — or merely produces a number that the clinician has to translate.

The Training Gap

There is also an educational dimension that the field is beginning to take seriously.

The reason advanced imaging-based selection is the central conversation in stroke right now is precisely because the evidence has outpaced the training.

Many frontline neurologists were not trained on the imaging-based mismatch paradigms that now drive selection.

Departments, fellowship programs, and continuing education infrastructure are now catching up to a paradigm that has been the trial standard for several years and is now becoming the practice standard.

The gap is real, and it sits between the centers that have already absorbed advanced imaging-based decision-making and the centers that have not.

Each evidence expansion in stroke intervention raises the imaging selection bar. Each rise in that bar widens the gap between trial conditions and the hospitals that have to implement them.

That gap is the adoption problem worth solving.

What Comes Next

DUSK, the ongoing Western complementary trial with Dr. Nogueira as PI at UPMC, will test whether ORIENTAL-MeVO‘s findings generalize beyond Chinese populations.

If positive, it closes the loop on Western validation. If negative, the field faces a real puzzle.

The 2026 AHA/ASA stroke guidelines, designed as a living-document architecture, are built to absorb evolving evidence rather than wait for the next full revision cycle.

The MeVO findings will shape subsequent updates.

The MeVO reversal is one chapter in a longer story the field is still writing.

The pattern is familiar: broad evidence followed by sharper selection, followed by the slow work of implementation. It has played out before and will play out again.

The harder work is downstream of the trials. It is the work of translating evolving selection criteria into reproducible bedside decisions across hospitals that look nothing like the trial sites.

That is where the field’s next decade will be spent. It is also where the gap between what the evidence shows and what the system delivers will be closed — or not.

References

Goyal M, et al. Endovascular treatment of stroke due to medium-vessel occlusion. NEJM 2025;392:1385-95.

Psychogios M, et al. Endovascular treatment for stroke due to occlusion of medium or distal vessels. NEJM 2025;392:1374-84.

Clarençon F, et al. Evaluation of mechanical thrombectomy in acute ischemic stroke related to a distal arterial occlusion. Int J Stroke 2024;19:367-72.

Hu W, et al. Endovascular treatment of medium-vessel-occlusion strokes. NEJM 2026;394:1894-904.

Saver JL, Gupta R, et al. DISTALS Trial: TIGERTRIEVER 13 in medium vessel occlusion stroke. International Stroke Conference 2026 late-breaking session.

Nogueira RG, et al. Distal medium vessel occlusion strokes: understanding the present and paving the way for a better future. J Stroke 2024;26:190-202.

Prabhakaran S, et al. 2026 AHA/ASA guideline for the early management of patients with acute ischemic stroke. Stroke, January 2026.

Nogueira RG, et al. Thrombectomy 6 to 24 hours after stroke with a mismatch between deficit and infarct (DAWN). NEJM 2018;378:11-21.

Albers GW, et al. Thrombectomy for stroke at 6 to 16 hours with selection by perfusion imaging (DEFUSE 3). NEJM 2018;378:708-18.”

Stay updated with Hemostasis Today.