Worse Stroke Outcome in Patients with Atrial Fibrillation May Be Due to Greater Volumes of More Severe Hypoperfusion

  • Dr Hans Tu, Department of Neurology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Dr Bruce Campbell, Department of Neurology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Dr Soren Christensen, Department of Radiology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Ms Marnie Collins, Department of Mathematics and Statistics, University of Melbourne, Australia
  • A/Prof Mark Parsons, Department of Neurology and Hunter Medical Research Institute, John Hunter Hospital, University of Newcastle, Australia
  • A/Prof P Alan Barber, Department of Medicine, University of Auckland, New Zealand
  • Prof Geoffrey Donnan, Florey Neuroscience Institutes, Australia
  • Prof Stephen Davis, Department of Neurology, Royal Melbourne Hospital, University of Melbourne, Australia

Background: Atrial fibrillation (AF) is associated with worse outcome following ischemic stroke, but the reasons for this remain unclear. We aimed to elucidate the pathophysiological determinants of worse stroke outcome in patients with AF using serial MRI data from the EPITHET study.
Methods: Comparisons of infarct size, hypoperfusion volume, infarct growth, recanalisation, reperfusion, hemorrhage, stroke severity, and response to thrombolysis were made between patients with and without AF in the EPITHET study.
Results: AF was present in 42 of 101 patients. At baseline, AF patients were older (79 vs 73 years, p=0.02), had more severe neurological impairment (NIHSS 16 vs 11, p=0.006), larger infarcts (29 vs 15mL p=0.04) and greater volumes of more severe hypoperfusion (Tmax≥8 Perfusion Weighted Image volume 70 vs 43mL, p=0.01) compared to patients without AF. There were no differences in arterial occlusion site, infarct growth, recanalisation or reperfusion. At outcome, AF patients had larger infarcts (52 vs 16mL, p=0.05), worse hemorrhagic transformation (29% vs 5%, p=0.002 for parenchymal hematomas) and higher mortality rates (31% vs 12%, p=0.04). AF was an independent predictor of parenchymal hematoma (OR 6.41, CI 1.53-26.81), but not mortality (OR 2.38, CI 0.79-7.10), after adjusting for baseline imbalances in age, tPA treatment, blood glucose, stroke size and severity.
Conclusions: Patients with AF have worse clinical and imaging outcomes following ischemic stroke. This study suggests that the adverse effect of AF is due to greater volumes of more severely hypoperfused tissue, leading to larger infarct size and greater risk of severe hemorrhagic transformation.