Thrombectomy beat drugs alone for mild large-vessel stroke in a China trial

A large artery in the brain can be blocked and yet produce only mild symptoms — a slightly weak hand, a little slurred speech — because, for a while, neighbouring vessels keep the threatened tissue alive. Those patients sit in an uncomfortable grey zone. Thrombectomy is proven for severe large-vessel strokes, but pushing a catheter into the brain of someone who is barely impaired risks turning a mild stroke into a disabling one. So most have been treated with drugs alone and watched [s1].

MILD-MT, published in the New England Journal of Medicine on 23 October, set out to test whether that caution is justified [s1]. It asked a direct question: for mild stroke from a large-vessel occlusion, does pulling the clot beat medication alone?

What the trial did

This was a multicentre, open-label, randomised trial run in China [s1]. It enrolled patients with an acute ischaemic stroke of mild clinical severity — defined as a score below 6 on the National Institutes of Health Stroke Scale — caused by a large-vessel occlusion of the anterior circulation [s1]. Eligibility also required a specific imaging picture: a small infarct core, meaning little tissue already dead, together with a large perfusion-mismatch volume, meaning a lot of tissue still at risk [s1]. Patients had to be treated within 24 hours of symptom onset or of when they were last known to be well [s1].

They were randomly assigned to endovascular therapy plus medical management, or to medical management alone [s1]. The primary outcome was an excellent functional outcome at 90 days, defined as a modified Rankin scale score of 0 — no symptoms — or 1 — no clinically significant disability [s1].

A total of 304 patients underwent randomisation, of whom 300 formed the modified intention-to-treat population: 144 in the endovascular group and 156 in the medical-management group [s1]. The trial is registered as MILD-MT, NCT06179017, and is now completed [s1] [s2].

What happened

At 90 days, an excellent outcome — a modified Rankin score of 0 or 1 — was recorded in 69.4% of the endovascular group and in 50.0% of the medical-management group (odds ratio, 2.32; 95% confidence interval, 1.41 to 3.81; P<0.001) [s1]. That is roughly a 19-percentage-point absolute gap in favour of clot removal, and the confidence interval sits well clear of no effect [s1].

It is worth being precise about what that endpoint means. This trial did not measure mere survival or basic independence; it measured a near-complete recovery, people walking away with little or no lasting disability [s1]. On that demanding bar, the procedure nearly doubled the odds of success [s1].

The catch: procedural complications

The safety picture has two very different halves [s1]. On bleeding, the usual worry with opening a brain artery, the trial was reassuring: symptomatic intracranial haemorrhage within 48 hours occurred in just 3 patients in each group [s1]. Among all participants, there was a single death, in the endovascular group [s1].

But the procedure itself was not gentle. Serious procedure-related complications occurred in 19 patients — 13.2% — in the endovascular group [s1]. That is roughly one in eight, and it is the figure that keeps this from being a simple, clean win. The benefit on function was large, the bleeding risk was not raised, and yet a meaningful fraction of patients paid a price at the catheter [s1].

How far the result travels

MILD-MT was conducted entirely in China, at centres able to select patients by perfusion imaging — a small core and a large mismatch [s1]. As with any single-country, imaging-selected trial, whether the same gap appears in health systems with different imaging access and different procedural experience is an open question [s1]. The 13.2% rate of serious procedural complications, in particular, is likely to depend on operator volume and technique, and may not transfer intact [s1].

What to watch

For years the default in mild large-vessel stroke has been to treat with drugs and intervene only if the patient deteriorates [s1]. MILD-MT is the kind of result that pushes against that default: among carefully imaged patients, early thrombectomy produced far more near-complete recoveries without raising brain bleeds [s1]. The unresolved question guideline writers now face is the procedural-harm rate — whether a one-in-eight risk of a serious complication is acceptable in patients who, by definition, started only mildly affected, and in which hands [s1].

This article describes trial results. It is not advice about any treatment, and nothing here should be used to start, stop or change care.