Opens in a new tab

Late Thrombectomy Extends Benefit and Risk

Thrombectomy improved the overall distribution of 90-day disability outcomes in selected patients treated 24 to 72 hours after a major stroke. Intracranial bleeding and death were also more frequent in the intervention group.

November 18, 2026
Editorial
A later arrival does not always mean there is no brain tissue left to save.PeopleImages / Shutterstock.com

IPM Take

LATE-MT adds evidence that the treatment window for some large-vessel strokes may depend on tissue condition as well as elapsed time. Its result is consequential precisely because it is uncomfortable: better functional outcomes came with more symptomatic bleeding and a higher observed death rate.

Executive Summary

The Chinese randomized trial assigned 336 patients with an anterior-circulation large-vessel occlusion and imaging-identified salvageable tissue to thrombectomy plus standard care or standard care alone. Patients were treated within a 24-to-72-hour window; the median time to randomization was 38.5 hours. The distribution of disability scores at 90 days favoured thrombectomy, with a common odds ratio of 1.74 for a better outcome. Symptomatic intracranial haemorrhage occurred in 4.8% of intervention patients and none of the controls; 90-day case fatality was 6.0% and 1.8%, respectively.

Why it matters

  • Patients and families: A late arrival may still warrant assessment, but the possible functional gain has to be discussed alongside bleeding and mortality risks.
  • Clinicians: Advanced imaging and specialist judgment are essential to identifying patients who resemble those enrolled in the trial.
  • Hospitals / providers: A later window would require dependable imaging, transfer and intervention capacity beyond the first day.
  • Payers: Planning for wider access should evaluate the full pathway and its outcomes, rather than counting additional procedures alone.

The LATE-MT trial tested whether patients arriving well after the first day of a large-vessel stroke could still benefit from reopening the artery. Eligibility depended on imaging that showed salvageable brain tissue. This was a highly selected population treated at comprehensive stroke centres, not a general sample of everyone presenting late.

The primary analysis found a more favourable distribution of modified Rankin Scale scores at 90 days after thrombectomy. It measured improvement across the range of disability outcomes rather than relying on one threshold for independence. Yet the safety results moved in the other direction: symptomatic intracranial haemorrhage occurred only in the intervention arm, and the observed death rate was higher there too.

The study makes late-window decisions more precise and more demanding. A service considering thrombectomy beyond 24 hours needs the ability to identify salvageable tissue, estimate procedural risk and discuss both with patients or families. Replication and guideline review will determine how this finding fits into routine stroke pathways; the 72-hour boundary should not be treated as a blanket instruction to intervene.

Source & Evidence