Skip to main content
NeuroWiki
Thrombectomy

DEFUSE 3 Trial: Thrombectomy for Ischemic Stroke (6-16 Hours)

In anterior LVO patients 6–16 hours from last known well with RAPID-defined perfusion mismatch, does endovascular thrombectomy improve the mRS distribution at 90 days compared with standard medical therapy?

Albers et al. (NEJM 2018;378(8):708–718) · doi:10.1056/NEJMoa1713973 · 182 patients

Population

Included

  • Anterior LVO (ICA or M1)
  • 6–16 hours from last known well
  • Pre-stroke mRS 0–1
  • RAPID perfusion mismatch: core <70 mL, mismatch ratio ≥1.8, mismatch volume ≥15 mL

Excluded

  • Onset <6 hours (use early-window evidence)
  • Onset >16 hours (use DAWN window)
  • Pre-stroke mRS ≥2
  • Posterior circulation occlusion (see ATTENTION/BAOCHE)
  • Core ≥70 mL or mismatch ratio <1.8

Source: ClinicalTrials.gov NCT02586415· Retrieved 2026-06-08

Primary Outcome — Ordinal mRS Shift

Ordinal-shift design: Primary endpoint is the full mRS distribution analyzed by ordinal logistic regression (common OR 2.77, 95% CI 1.63–4.70). The mRS 0–2 dichotomization shown below is a secondary outcome; its NNT 3.6 is derived from the secondary, not the primary.

Thrombectomy + MedicalBetter outcome
45 / 100
Medical Therapy Alone
17 / 100

mRS 0–2 at 90 Days (secondary)

Risk ratio cOR 2.7795% CI 1.63–4.70p = <0.001

Study Arms

Agent
Thrombectomy with any FDA-approved thrombectomy device, at the neurointerventionalist's discretion. Carotid angioplasty ± stenting permitted for cervical ICA atherosclerotic stenosis/occlusion.
Route
Endovascular
Frequency
Single procedure, 6–16 h from last-known-well. Protocol required femoral puncture within 90 minutes after the end of qualifying imaging. General anesthesia discouraged. Intra-arterial t-PA not allowed.
Duration
One-time procedure
Co-interventions
Standard medical therapy per current AHA guidelines (BOTH groups). IV t-PA allowed only if begun within 4.5 h of onset.

Technique, 90-min puncture target, GA-discouraged, no-IA-tPA, and AHA-guideline medical therapy in both groups from Albers NEJM 2018 p.710. Technical success = TICI 2b/3.

Safety

Symptomatic ICH

Thrombectomy + Medical

7%

Medical Therapy Alone

4%

7% EVT vs 4% control (P=0.75). No significant difference.

90-day mortality

Thrombectomy + Medical

14%

Medical Therapy Alone

26%

14% EVT vs 26% control (P=0.05). Borderline significant secondary outcome. Directional benefit.

Trial Design

Type

  • Multicenter randomized open-label trial
  • Blinded-endpoint assessment
  • Perfusion imaging selection (CTP/MRI)
  • 1:1 allocation (Thrombectomy vs. Medical)

Timeline

Enrolled 2016-2017

N

182

Enrollment

182 patients (92 EVT / 90 medical) at 38 US centers. Enrolled May 2016 – Dec 2017. Stopped early at pre-specified interim. Published NEJM 2018.

ClinicalTrials.gov

NCT02586415

Bedside Pearl

In anterior LVO 6–16 hours from LKW with RAPID-defined penumbra (core <70 mL, mismatch ≥1.8 and ≥15 mL), EVT triples functional independence (45% vs 17%). Primary endpoint is the ordinal mRS shift (common OR 2.77, CI 1.63–4.70); NNT 3.6 derived from the mRS 0–2 secondary.

NeuroWiki is a clinical reference. It does not substitute for your clinical judgment, current guidelines, or your institution's protocol. Verify before acting. Do not enter patient names, MRNs, or dates of birth. Privacy Policy