Prognostic Significance of Hyperglycemia in Patients with Brain Tumors: a Meta-Analysis.

Abstract

Hyperglycemia has been associated with poor outcomes of patients with various diseases. There were several studies published to assess the association between hyperglycemia and prognosis of patients with brain tumors, but no consistent conclusion was available. We therefore performed a meta-analysis of available studies to evaluate the prognostic role of hyperglycemia in brain tumors. Several common databases were searched for eligible studies on the association between hyperglycemia and survival of patients with brain tumors. Two investigators used a set of predefined inclusion criteria to assess eligible studies independently. The pooled hazard ratios (HRs) with 95% confidence intervals (95% CIs) were used to assess the prognostic role of hyperglycemia. Finally, seven studies with a total of 2168 patients with brain tumors were included into the meta-analysis. Meta-analysis of total seven studies showed that hyperglycemia was significantly associated with shorter overall survival of brain tumors (HR = 2.04, 95% CI 1.51-2.76, P < 0.001). Meta-analysis of studies focusing on hyperglycemia showed that hyperglycemia was still significantly associated with shorter overall survival of brain tumors (HR = 1.82, 95% CI 1.29-2.59, P = 0.001). Meta-analysis of three studies on diabetes showed that diabetes was significantly associated with shorter overall survival of brain tumors (HR = 2.09, 95% CI 1.22-3.57, P = 0.007). Meta-regression analysis showed that there was no obvious difference in the roles of between hyperglycemia caused by glucocorticoids and hyperglycemia from diabetes (P = 0.25). Thus, hyperglycemia has an obvious prognostic significance in patients with brain tumors, and hyperglycemia is significantly associated with shorter overall survival of brain tumors.