Abstract 14: Receiver Operating Characteristics Curve Analysis Demonstrates That Quantitative Assessment Of Iodine Contrast Extravasation Post Reperfusion Therapy For Ischemic Stroke Predicts Reperfusion Injury And Discharge Modified Rankin

Abstract

Listen

Translate article

Introduction: We previously demonstrated that the quantitative volumetric assessment of iodinated contrast extravasation (ICE) present on post-intervention imaging was closely associated with the likelihood of an acute ischemic stroke patient having PH-1 or PH-2, and thus ICE may serve as a useful CT biomarker to assess risk of reperfusion injury (hemorrhagic conversion (HC) and blood brain barrier (BBB) disruption). Here we used receiver operator curve (ROC) analysis to compare the efficacy of ICE, infarct volume, and 24hr NIHSS change as a predictor of discharge mRS and HC post-reperfusion therapy. Method: Data on ischemic stroke patients treated with reperfusion therapy were obtained from our Institutional Review Board approved database from January 2017 to November 2019 that had evaluable images within 24 hours of admission. Ischemic volume (IV) was measured on diffusion-weighted imaging. ICE was measured on CT head. A freehand 3D region of interest tool on the Visage Imaging PACS System was used to measure volumes. Susceptibility weighted MRI sequences were used to grade HC. Data analysis was conducted with regression modeling and ROC analysis. Results: Of the 82 patients, median age was 73 (interquartile range (IQR) 61- 77, 49% were women, admission NIHSS was 12 (IQR 7 - 21), 24hr NIHSS change was 4 (IQR 0 -13), IV was 50.6 +/- 7.1 mL, 48% were treated with thrombectomy, 7% had PH-1 or PH-2 identified on MRI, median systolic blood pressure was 154 (IQR 137-175), 56% were MCA territory strokes, and 37% had a discharge mRS of 0-2. ICE volume was 2.6 +/- 1.0 mL. ICE increased the likelihood of PH-1 or PH-2 HC (odds ratio (OR) 14.34, 95% confidence interval (CI) 5.74 - 22.94) and decreased the likelihood of discharge with mRS of 0-2, OR of 0.09 (CI 0.008-0.972). IV was a better predictor of 0-2 mRS (ROC area under the curve (AUC) 0.832) than ICE (AUC 0.640) and 24hr NIHSS change (AUC 0.557), but ICE was a better predictor of PH-1 or PH-2 (AUC 0.942) than IV (AUC 0.667) and 24hr NIHSS change (AUC 0.447). Conclusion: ICE may predict reperfusion injury and functional outcome, but it is a better predictor of hemorrhagic conversion in patients treated with reperfusion therapy.