Admission Hyperglycemia and Clinical Outcome in Cerebral Venous Thrombosis.

Abstract

Admission hyperglycemia is associated with poor clinical outcome in ischemic and hemorrhagic stroke. Admission hyperglycemia has not been investigated in patients with cerebral venous thrombosis. Consecutive adult patients with cerebral venous thrombosis were included at the Academic Medical Center, The Netherlands (2000-2014) and the Helsinki University Central Hospital, Finland (1998-2014). We excluded patients with known diabetes mellitus and patients without known admission blood glucose. We defined admission hyperglycemia as blood glucose ≥7.8 mmol/L (141 mg/dL) and severe hyperglycemia as blood glucose ≥11.1 mmol/L (200 mg/dL). We used logistic regression analysis to determine if admission hyperglycemia was associated with modified Rankin Scale (mRS) score of 3 to 6 or mortality at last follow-up. We adjusted for: age, sex, coma, malignancy, infection, intracerebral hemorrhage, deep cerebral venous thrombosis, and location of recruitment. Of 380 patients with cerebral venous thrombosis, 308 were eligible. Of these, 66 (21.4%) had admission hyperglycemia with 8 (2.6%) having severe admission hyperglycemia. Coma (31.3% versus 5.0%, P<0.001) and intracerebral hemorrhage (53.0% versus 32.6%, P=0.002) at presentation were more common among patients with admission hyperglycemia than normoglycemic patients. Patients with admission hyperglycemia had a higher risk of mRS score of 3 to 6 (adjusted odds ratio, 3.10; 95% confidence interval, 1.35-7.12) and mortality (adjusted odds ratio, 4.13; 95% confidence interval, 1.41-12.09). Severe hyperglycemia was even more strongly associated with mRS score of 3 to 6 (adjusted odds ratio, 11.59; 95% confidence interval, 1.74-77.30) and mortality (adjusted odds ratio, 33.36; 95% confidence interval, 3.87-287.28) compared with normoglycemic patients. Admission hyperglycemia is a strong predictor of poor clinical outcome in patients with cerebral venous thrombosis.