Effect of Hemoglobin and Blood Glucose Levels on CT Perfusion Ischemic Core Estimation

Background and objectives: CT perfusion (CTP) maps can estimate the ischemic core in acute ischemic stroke based on distinctive cerebral blood flow thresholds. However, metabolic factors beyond perfusion influence the tissue tolerance to ischemia and the infarct growth rate. Underestimating the ischemic core volume (ICV) might result in overestimating the salvageable cerebral tissue and, consequently, overestimating the potential clinical benefits of reperfusion therapies. We aim to evaluate whether baseline hemoglobin and blood glucose levels influence the accuracy of baseline CTP ICV estimations. Methods: Large vessel occlusion stroke patients investigated with baseline CTP undergoing thrombectomy with near-complete reperfusion and without parenchymal hemorrhage from the ESCAPE-NA1 trial were included. Patients were subdivided into anemic (hemoglobin <130 g/L for men and <120 g/L for women) and nonanemic groups, and hyperglycemic (blood glucose level >7 mmol/L) and normoglycemic groups. Ischemic core underestimated volume (ICuV) was calculated: final infarct volume minus CTP-based ICV. The primary outcome was the presence of "perfusion scotoma" defined as ICuV ≥10 mL. Presence of "perfusion scotoma" and median ICuV were compared between anemic vs nonanemic and hyperglycemic vs normoglycemic patients using nonparametric tests and multivariable binary logistic regression with adjustment for baseline variables. Results: One hundred sixty-two of 1,105 (15%) patients were included (median age 70.5 [interquartile range (IQR) 61-80.4], 50.6% women). The median ICuV was 7.26 mL (IQR 0-25.63). Seventy-eight (48%) patients demonstrated perfusion scotoma. Forty-two (25.7%) patients were anemic, and 65 (40.1%) were hyperglycemic. In univariable analysis, the hyperglycemic group had a higher prevalence of perfusion scotoma (65% [n = 40] vs 39% [n = 38], p = 0.006) and larger ICuV (17.79 mL [IQR 1.57-42.75] vs 6 mL [-0.31 to 12.51], p = 0.003) compared to normoglycemic patients. No significant ICuV differences between patients with and without anemia were seen. Multivariable regression analysis revealed an association between perfusion scotoma and hyperglycemia, adjusted odds ratio (OR) 2.48 (95% CI 1.25-4.92), and between perfusion scotoma and blood glucose levels, adjusted OR 1.19 (95% CI 1.03-1.39) per 1 mmol/L increase. Discussion: In our study, CTP-based ischemic core underestimation was common and associated with higher baseline blood glucose levels. Individual metabolic factors beyond perfusion that critically influence the infarct growth rate should be considered when interpreting baseline CTP estimations of ischemic core.