Abstract A031: Non-Contrast CT Severe Hypodensity versus Net Water Uptake in Assessing Thrombectomy Treatment Effect in Large Core Stroke: A Secondary Analysis of the SELECT2 Trial

Introduction: Severe CT-hypodensity, representing blood-brain-barrier (BBB) injury, has been shown to modify endovascular therapy (EVT) treatment effect in patients presenting with large core stroke. Net water uptake (NWU), reflecting average hypodensity within the lesion, is another measure of BBB injury and has been shown to be prognostic of poor outcome, but whether it modifies EVT treatment effect is unclear. Methods: We performed a comparative analysis of severe CT-hypodensity and NWU in patients enrolled into SELECT2, a randomized trial of EVT and medical management (MM) in patients with large core stroke (ASPECTS 3-5 or CTP core ≥50ml). Visible CT-hypodensity was manually outlined on baseline imaging. Severe CT-hypodensity was defined as the volume of core voxels <26HU (Hounsfield units). NWU was calculated using the standard method of percentage relative hypodensity versus a contralateral mirror region-of-interest, after exclusion of CSF and artifacts. The primary outcome was mRS 0-3 at 90 days. The associations between severe CT-hypodensity, NWU and the primary outcome were evaluated in separate logistic regression models within EVT and MM arms, adjusted for age, NIHSS, and total CT-core volume. Model performance between severe hypodensity and NWU was compared using Bayesian Information Criterion (BIC). Results: In 318 patients (51% EVT, 49% MM), the median volume of severe hypodensity (EVT: 13.8mL IQR: 5.1 - 27.0, MM: 15.9mL IQR: 6.7 - 31.5) and NWU (EVT: 9.2% IQR: 6.3 - 13.3, MM: 9.8% IQR: 7.4 - 13.4) was balanced between the two treatment groups. The volume of severe hypodensity (per 1mL increase) was associated with a lower odds of mRS 0-3 with EVT (aOR=0.97, 95% CI:0.94-99) but not with MM (aOR=1.01, 95% CI:0.98-1.05, p-interaction<0.01). An increase in NWU (per % increase) was not significantly associated with a lower odds of mRS 0-3 with EVT (aOR=0.96, 95%CI: 0.89-1.024) or MM (aOR=0.96, 95% CI:0.87-1.05, p-interaction=0.86; Figure). Models including severe hypodensity provided a better fit for the primary outcome, compared to models including NWU (delta BIC of 5, favoring severe hypodensity). Conclusions: In patients with large core stroke, the volume of severe CT-hypodensity was associated with clinical outcome and modified EVT treatment effect. NWU did not modify EVT treatment effect. These findings suggest that severe CT-hypodensity may be the preferred imaging criterion when considering EVT treatment decisions.