Associations of anticoagulant use with outcome in newly diagnosed glioblastoma

BACKGROUND To test the hypothesis that despite bleeding risk, anticoagulants improve the outcome in glioblastoma because of reduced incidence of venous thromboembolic events and modulation of angiogenesis, infiltration and invasion. METHODS We assessed survival associations of anticoagulant use from baseline up to the start of temozolomide chemoradiotherapy (TMZ/RT) (period I) and from there to the start of maintenance TMZ chemotherapy (period II) by pooling data of three randomised clinical trials in newly diagnosed glioblastoma including 1273 patients. Progression-free survival (PFS) and overall survival (OS) were compared between patients with anticoagulant use versus no use; therapeutic versus prophylactic versus no use; different durations of anticoagulant use versus no use; anticoagulant use versus use of anti-platelet agents versus neither anticoagulant nor anti-platelet agent use. Cox regression models were stratified by trial and adjusted for baseline prognostic factors. RESULTS Anticoagulant use was documented in 75 patients (5.9%) in period I and in 104 patients (10.2%) in period II. Anticoagulant use during period II, but not period I, was associated with inferior OS than no use on multivariate analysis (p = 0.001, hazard ratio [HR] = 1.52, 95% confidence interval [CI]: 1.18-1.95). No decrease in OS became apparent when only patients with prophylactic anticoagulant use were considered. No survival association was established for anti-platelet agent use. CONCLUSIONS Anticoagulant use was not associated with improved OS. Anticoagulants may not exert relevant anti-tumour properties in glioblastoma