Which glucose parameter best predicts poor outcome after mechanical thrombectomy for acute large vessel occlusion stroke?

Background Hyperglycaemia is related to adverse outcomes in patients with acute ischaemic stroke (AIS) treated with mechanical thrombectomy (MT). Aims To compare the predictive ability of admission blood glucose (ABG), fasting blood glucose (FBG) and glycated haemoglobin (HbA1c) in patients with acute large vessel occlusion (ALVO) stroke undergoing MT. Methods The study retrospectively analysed the data of 329 consecutive acute stroke patients treated with MT. Multivariate logistic regression analysis was performed to investigate the predictors of poor functional outcome. Receiver operating characteristic (ROC) analysis was performed to identify the predictive ability of ABG, FBG and HbA1c for poor functional independence in patients with ALVO stroke undergoing MT. In addition, subgroup analyses were performed for both diabetic and non-diabetic patients. Results Multiple logistic regression analysis demonstrated that ABG and FBG were independent predictors of 3-month poor functional outcome; however, HbA1c could not predict poor functional outcome. ROC analysis showed that FBG has a higher predictive ability than ABG (areas under the ROC curve (AUC): 0.689 vs 0.624; P = 0.037) and HbA1c (AUC: 0.689 vs 0.541; P < 0.001) for poor function outcome. Conclusions ABG and FBG are independent predictors of poor functional outcome in patients with AIS undergoing MT, and FBG has a higher predictive ability than ABG and HbA1c.