DDEL-12. Superselective intra-arterial (IA) bevacizumab administration decreases venous thromboembolism (VTE) incidence compared to standard intravenous (IV) bevacizumab in glioblastoma multiforme (GBM)

Abstract BACKGROUND Bevacizumab, a monoclonal antibody targeting vascular endothelial growth factor (VEGF) has been associated with an increased risk of venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE). While intravenous (IV) administration is the standard, intra-arterial (IA) delivery has emerged as an alternative and may reduce systemic side effects. This study compares VTE incidence in glioblastoma (GBM) patients receiving IA versus IV bevacizumab. METHODS A retrospective analysis was conducted on 133 patients with newly diagnosed or recurrent GBM who received bevacizumab: IA-alone (n=58), IV-alone (n=61), or both IA and IV (n=14). For seven patients who transitioned from IA to IV, VTEs were attributed to IA if they occurred within 30 days of the last IA dose, and to IV if > 30 days of last IA dose. RESULTS In the IA-alone group (n=58/133), DVT occurred in 3.4% (2/58), PE in 1.7% (1/58), and combined DVT+PE in 5.2% (3/58). In the IV-alone group (n=61/133), DVT occurred in 23.0% (14/61), PE in 4.9% (3/61), and DVT+PE in 9.8% (6/61); this group had a significantly higher DVT incidence than IA-alone (23.0% vs.3.4%, p = 0.0016). The IA+IV group (n=14/133) had a DVT rate of 21.4% (3/14) and DVT+PE rate of 7.1% (1/14). All 4 VTE events in IA+IV occurred after IV bevacizumab. All VTEs in seven patients switched from IA to IV were attributed to IV. Any VTE occurred in 37.7% of IV-alone and 10.3% of IA-alone patients (p = 0.015). CONCLUSION This study suggests an increased VTE incidence for IV versus IA bevacizumab administration. IA delivery may reduce systemic exposure, potentially lowering thrombotic risk, aligning with the lower VTE rates observed in the IA-alone group. Given the clinical significance of VTE, further research is needed to confirm these findings in larger cohorts.