Abstract WMP91: Ischemic Injury Extent on Different Imaging Modalities and Its Association With Endovascular Thrombectomy Outcomes for Large Core Stroke: A SELECT2 Imaging Analysis

Abstract

Introduction: The association of extent of ischemic injury on various imaging modalities and EVT efficacy and safety in patients with large ischemic core remains unexplored. We analyzed ischemic injury estimates on structural and perfusion imaging modalities and their association with time and EVT treatment effect. Methods: In SELECT2, all patients received non contrast CT and CT perfusion/MR diffusion. Baseline ischemic injury was estimated using ASPECTS, CTP/MRI with RAPID processing, manual delineation of CT hypodensity and composite core [the larger of the CT hypodensity and CTP/MRI core volumes], Figure 1A. We assessed how these estimates compared to one another, which correlated best with the outcomes and described EVT treatment effect across their strata. Results: Of 352 patients, 16 were excluded for missing mRS/imaging data. 170/336 (51%) received EVT. The median (IQR) CT-ASPECTS was 4 (3-5), CT-hypodensity 86 (49-114) mL, CTP/MRI core 73 mL (46-107). 60% had CT hypodensity > CTP/MRI core volume. CTP core was larger in 81% within 0-3 hours that inverted to 86% patients with larger CT hypodensity in 21-24 hours of LKW [Figure 1B]. Composite core (101 [72-138] ml) had best fit for mRS (Bayesian Information Criteria for mRS shift: ASPECTS - 448, CT hypodensity - 443, CTP core - 434, Composite core - 429 with lower the better). Treatment effect estimates favored EVT across strata (≥70 ml, ≥100ml & ≥150 ml) for CT hypodensity, CTP/MRI core and composite core as well as ASPECTS 3,4 and 5. For a given volume probability of independent ambulation with EVT decreased with age and time to reperfusion. Conclusions: CT perfusion and CT hypodensity were complementary and most prognostic when used together, in conjunction with age and time to reperfusion. Thrombectomy benefit was preserved across ischemic volumes and ASPECTS. These findings can assist clinicians in assessing the likely outcome of thrombectomy for individual patients. Trial Registration: NCT03876457