Acute stroke: doubt cast on benefit of prethrombectomy IV alteplase

  • Yang P & al.
  • N Engl J Med
  • 6 May 2020

  • curated by Susan London
  • Clinical Essentials
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.


  • Patients with large-vessel occlusion acute ischemic stroke had a noninferior neurologic outcome when they underwent endovascular thrombectomy directly, without receiving intravenous (IV) alteplase first.

Why this matters

  • Alteplase is costly and carries bleeding risk.
  • Editorial points to a "generous" margin for noninferiority, says the confidence intervals did not exclude a benefit of ~20% with combination therapy, and urges continued adherence to current guidelines recommending prethrombectomy alteplase in all eligible patients.

Key results

  • Vs thrombectomy preceded by IV alteplase, thrombectomy alone yielded noninferior distribution of modified Rankin scale scores at 90 days:
    • Median score (interquartile range) of 3 (2-5) for each;
    • Adjusted common OR, 1.07 (95% CI, 0.81-1.40); and
    • P=.04 for noninferiority.
  • However, thrombectomy alone netted lower rates of:
    • Successful reperfusion before thrombectomy (2.4% vs 7.0%) and
    • Successful reperfusion overall (79.4% vs 84.5%).
  • Groups were similar on 90-day mortality (17.7% vs 18.8%; P=.71), other serious adverse events (e.g., intracranial hemorrhage, infarction in new territory).

Study design

  • Multicenter noninferiority randomized controlled trial among 656 patients in China with acute ischemic stroke from large-vessel occlusion in anterior circulation (DIRECT-MT trial).
  • Randomization: endovascular thrombectomy with vs without preprocedural IV alteplase (0.9 mg/kg; ≤4.5 hours after symptom onset).
  • Main outcome: distribution of modified Rankin scale scores (noninferiority: lower boundary of 95% CI of adjusted common OR ≥0.8).
  • Funding: Chinese government; Wu Jieping Medical Foundation.


  • Protocols followed 2015 guidelines.
  • Larger noninferiority margin, smaller sample.
  • Region-specific triage system, consent process.