Abstract 159: Baseline Perfusion Imaging Collateral Scores Predict Infarct Growth in DEFUSE 3

Abstract

Background: High relative cerebral blood volume (rCBV) within the ischemic lesion and a low hypoperfusion intensity ratio (HIR) predict good collateral vessel status and correlate with infarct growth and functional outcome in early window patients with large vessel anterior circulation occlusions. Their performance in predicting clinical and radiologic outcome has not been assessed in late presenting patients. We hypothesized that favorable baseline HIR and rCBV profiles would predict less infarct growth at 24 hours in late presenting (6-16h), non-reperfused patients in DEFUSE 3. Methods: Non-reperfused patients in both arms of DEFUSE 3 were included. Baseline ischemic core, Tmax >6s, and Tmax >10s perfusion lesion volumes were calculated with RAPID software; 24h infarct volumes were manually determined from DWI or CT. Substantial infarct growth was defined as a >25mL increase between baseline ischemic core and 24h infarct volume. HIR was defined as the proportion of the Tmax >6s lesion with Tmax >10s delay; rCBV was calculated by RAPID from the mean CBV values within the Tmax >6s lesion compared to regions of normal CBV. Functional independence was defined as a modified Rankin score of 0-2. Results: Eighty-four patients with baseline perfusion imaging were included. ROC analysis showed that HIR >0.34 (AUC=0.68) and rCBV <0.74 (AUC=0.72) optimally predicted substantial infarct growth at 24h. Median growth was 23.4 versus 73.2mL with the HIR threshold of 0.34 (p=0.005), and 24.3 versus 58.7mL with the rCBV threshold of 0.74 (p=0.004). If baseline HIR and rCBV were both favorable, median growth was 21.7mL, 40.9mL if one was favorable, and 108.2mL if both were unfavorable (p=<0.001). Baseline HIR and rCBV were not associated with functional outcome at 90 days, though baseline Tmax >10s volume was significantly smaller in functionally independent patients (p=0.04). Conclusion: Perfusion collateral scores (HIR and rCBV) forecast infarct growth in late presenting, non-reperfused ischemic stroke patients. These parameters may be useful for guiding transfer decisions, such as need for repeat imaging on arrival to a thrombectomy center, and may help identify slow progressing patients more likely to have persistent salvageable ischemic tissue beyond 24 hours.