Microbleeds: increased recurrence risk in embolic stroke of undetermined source

  • Shoamanesh A & al.
  • JAMA Neurol
  • 19/10/2020

  • Susan London
  • Clinical Essentials
L'accesso ai contenuti di questo sito è riservato agli operatori del settore sanitario italiano L'accesso ai contenuti di questo sito è riservato agli operatori del settore sanitario italiano

Takeaway

  • Risk for another stroke is 51% higher in patients with embolic stroke of undetermined source who have cerebral microbleeds.
  • The presence of cerebral microbleeds does not significantly affect rivaroxaban efficacy and safety.

Why this matters

  • Editorial: findings should “contribute to changing the mindset of clinicians," and "[c]urrent evidence does not justify avoiding antithrombotic medication" solely because of microbleeds.

Key results

  • 11% of patients had microbleeds.
  • Independent predictors (ORs):
    • Advancing age (OR per year): 1.03.
    • East Asian ancestry: 1.57.
    • Hypertension: 2.20.
    • Multiterritorial infarcts: 1.95.
    • Chronic infarcts: 1.78.
    • Occult intracerebral hemorrhage: 5.23.
  • Microbleeds increased HRs (95% CIs) for:
    • Recurrent stroke: 1.51 (1.02-2.25).
    • Intracerebral hemorrhage: 4.18 (1.26-13.90).
    • All-cause mortality: 2.13 (1.06-4.26).
  • Lobar microbleeds specifically increased risk for ischemic stroke: HR, 2.33 (95% CI, 1.26-4.30).
  • No significant interactions between microbleeds and treatment (rivaroxaban vs aspirin) for recurrent stroke, ischemic stroke, or all-cause mortality risks.
  • Intracerebral hemorrhage risk on rivaroxaban was similar with (HR, 3.12) and without microbleeds (HR, 2.96; interaction P>.99).

Study design

  • Subgroup analyses of international, phase 3, randomized clinical trial comparing rivaroxaban vs aspirin in patients aged ≥50 years with embolic stroke of undetermined source (NAVIGATE ESUS).
  • 3699 patients (51%) had information on cerebral microbleeds reported on baseline MRI.
  • Main outcome: recurrent stroke over a median 11 months.
  • Funding: Bayer AG; Janssen Research and Development LLC.

Limitations

  • Possible selection bias, misclassification.
  • Lack of power to assess associations of microbleed burden, topography with outcomes.