Abstract WMP24: Collateral Status Contributes to Differences Between Observed and Predicted 24-Hour Infarct Volumes in DEFUSE 3

Abstract

Background and Purpose: We have previously shown that in the DEFUSE 3 trial, the infarct volume 24 hours after randomization was predicted by the union of the baseline core and the 24-hour Tmax>6s perfusion lesion. We determined if collateral robustness measured by the hypoperfusion intensity ratio (HIR) and cerebral blood volume index (CBV) accounts for the variance in infarct volume predictions. Methods: DEFUSE 3 patients underwent MRI with perfusion or CT perfusion at baseline and 24 hours after randomization. We used RAPID software to determine ischemic core and Tmax>6s lesion volumes as well as HIR and CBV Index at baseline and 24 hours. Patients were stratified by the difference between the predicted and the observed infarct volume at 24 hours. We compared baseline and follow-up HIR and CBV Index, as well as several other imaging and clinical outcomes in subgroups based on the accuracy of the infarct volume estimate. Results: Out of 123 eligible patients, 34 had 24-hour infarcts larger than predicted and these patients had less favorable collaterals (HIR 0.43 vs 0.32, p=0.006 and CBV Index 0.78 vs 0.85, p=0.001) at baseline, as well as at 24-hour follow-up (HIR 0.56 vs 0.07, p=0.004 and CBV Index 0.47 vs 0.73, p=0.006) compared to the 71 patients with more accurate infarct volume prediction. The remaining 18 patients had 24-hour infarct volumes smaller than predicted. These patients had similar baseline collateral scores but a more favorable CBV Index at 24 hours (0.81 vs 0.73, p=0.040) compared to patients with more accurate infarct prediction. Conclusions: Patients with 24-hour infarcts larger than predicted had evidence of less favorable baseline collaterals that fail within 24 hours, while patients with 24-hour infarcts smaller than predicted typically had favorable collaterals that persisted for at least 24 hours. Clinical Trial Registration: URL: http://www.clinicaltrials.gov. Unique identifier: NCT02586415