The effect of hyperglycemia on survival in patients with glioblastoma multiforme

Abstract

2013 Background: Steroid-induced hyperglycemia is rarely carefully managed in patients with glioblastoma multiforme (GBM) unless acute complications of hyperglycemia are present; however epidemiologic and laboratory studies suggest that hyperinsulinemia and/or hyperglycemia may promote tumor growth. We aimed to evaluate whether hyperglycemia was associated with decreased survival in patients with GBM. Methods: 191 patients with newly diagnosed GBM were accrued to New Approaches to Brain Tumor Therapy CNS Consortium trials with similar eligibility criteria between 1999 and 2005, prior to the standard use of temozolomide. 123 patients (64%) had glucose values and glucocorticoid doses recorded every 2 months for at least 6 months after accrual or until death. For each subject, mean glucose and mean glucocorticoid dose were calculated from all available results after weighting each value according to the number of days between it and the next record or the censor date. The primary outcome was survival. Results: The mean age was 55.7 years (SD 11.2 years) and the mean Karnofsky performance score at the time of accrual was 87 (SD 10.2). The median of subjects’ average glucocorticoid dose was 5 mg/day (range 0–64 mg) of dexamethasone or the equivalent. The median average glucose was 110 mg/dl (range 65–459 mg/dl), and 41 patients (21%) had a mean glucose > 140 mg/dl (the cut-off for impaired glucose tolerance). The median survival was longer in subjects with a mean glucose < 140 mg/dl than > 140 mg/dl (372 days versus 230 days, p=0.005). After adjusting for age, sex, race, baseline Karnofsky performance status, and mean glucocorticoid dose, every 10 mg/dl increase in mean glucose was associated with a 3% increase in the risk of death (95% CI: 0%-6%, p=0.034). Conclusions: In a group of patients with GBM and a good baseline performance status, hyperglycemia was associated with shorter survival after controlling for measured confounders. Interventional studies are needed to determine whether intensive management of glucocorticoid-related hyperglycemia in patients with GBM improves survival. No significant financial relationships to disclose.