Final Infarct Volume Can Be Determined Using Diffusion Imaging 24 Hours After Stroke Onset

  • Dr Bruce Campbell, Department of Medicine and Neurology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Dr Soren Christensen, Department of Radiology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Dr Hans Tu, Department of Medicine and Neurology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Prof Patricia Desmond, Department of Radiology, Royal Melbourne Hospital, University of Melbourne, Australia
  • A/Prof Christopher Levi, Hunter New England Area Health Service, Australia
  • Prof Christopher Bladin, Australia
  • Dr Niels Hjort, Denmark
  • Dr Mahmoud Ashkanian, Denmark
  • Dr Christine Solling, Denmark
  • Prof Geoffrey Donnan, Florey Neuroscience Institutes, University of Melbourne, Australia
  • Prof Stephen Davis, Department of Medicine and Neurology, Royal Melbourne Hospital, University of Melbourne, Australia
  • Dr Leif Ostergaard, Center of Functionally Integrative Neuroscience, Department of Neuroradiology, Århus University Hospit, Denmark
  • A/Prof Mark Parsons, of Neurology and Hunter Medical Research Institute, John Hunter Hospital, University of Newcastle, Australia

Objective: Infarct growth is commonly used as a surrogate endpoint in clinical trials but is traditionally assessed at day 90. However, this timepoint underestimates infarct volume due to atrophy in the affected region and the delay increases loss to follow-up. We hypothesised that infarct evolution is complete by 24 hours and correlated 24 hour and day 90 infarct volumes using MRI.
Methods: Infarct volume was assessed using manually drawn regions-of-interest on diffusion-weighted-imaging (DWI) at baseline and 24 hours after stroke onset as well as on FLAIR at day 90.
Results: Imaging data were available for 77 patients, 71 treated with tPA. Mean infarct volume was 36mL (st dev 43mL) on 24hr DWI. The DWI infarct volume at 24 hours had strong linear correlation with day 90 FLAIR infarct volume (r=0.97, 95% CI 0.95-0.98). The baseline DWI-day 90 correlation (assessable in 40 patients) was significantly lower, r=0.91 (95% CI 0.84-0.95). Infarct growth occurred between baseline and 24hours (mean 23mL absolute, 240% relative with 17/40 (43%) patients more than doubling their infarct volume). There were no patients with expansion of infarction between 24hrs and day 90. In fact, significant infarct shrinkage occurred from 24 hours to day 90 (mean volume 28mL, st dev 40mL, paired Wilcoxon p<0.001).
Conclusions: Infarct growth occurs within the first 24 hours. By 24 hours, DWI lesion volume accurately reflects the final infarct volume at day 90. Use of an earlier assessment has the potential to expedite trials and reduce loss to follow-up.