Prediction of adverse outcomes by blood glucose level after endovascular therapy for acute ischemic stroke

Object The authors evaluated the prognostic significance of blood glucose level at admission (BGA) and change in blood glucose at 48 hours from the baseline value (CG48) in nondiabetic and diabetic patients before and after endovascular therapy for acute ischemic stroke (AIS). Methods The BGA and CG48 data were analyzed in 614 patients with AIS who received endovascular therapy at 7 US centers between 2006 and 2009. Data reviewed included demographics, stroke risk factors, diabetic status, National Institutes of Health Stroke Scale (NIHSS) score at presentation, recanalization grade, intracranial hemorrhage (ICH) rate, and 90-day outcomes (mortality rate and modified Rankin Scale score of 3–6 [defined as poor outcome]). Variables with p values < 0.2 in univariate analysis were included in a binary logistic regression model for independent predictors of 90-day outcomes. Results The mean patient age was 67.3 years, the median NIHSS score was 16, and 27% of patients had diabetes. In nondiabetic patients, BGA ≥ 116 mg/dl (≥ 6.4 mmol/L) and failure of glucose level to drop > 30 mg/dl (> 1.7 mmol/L) from the admission value were both significant predictors of 90-day poor outcome and death (p < 0.001). In patients with diabetes, BGA ≥ 116 mg/dl (≥ 6.4 mmol/L) was an independent predictor of poor outcome (p = 0.001). The CG48 was not a predictor of outcome in diabetic patients. A simplified 6-point scale including BGA, Thrombolysis in Myocardial Infarction (TIMI) Grade 2–3 Reperfusion, Age, presentation NIHSS score, CG48, and symptomatic ICH (BRANCH) corresponded with poor outcomes at 90 days; the area under the curve value was > 0.79. Conclusions Failure of blood glucose values to decrease in the first 48 hours after AIS intervention correlated with poor 90-day outcomes in nondiabetic patients. The BRANCH scale shows promise as a simple prognostication tool after endovascular therapy for AIS, and it merits prospective validation.