Single-institution experience using bevacizumab for <i>IDH1/2</i> -mutant gliomas.

Abstract

e14031 Background: Bevacizumab (Bev), a monoclonal antibody that inhibits vascular endothelial growth factor, was fully FDA-approved in 2017 for use in recurrent glioblastoma and is often considered and utilized to treat all recurrent gliomas. Currently, there are limited studies on Bev use for recurrent isocitrate dehydrogenase 1/2 mutant ( IDH MUT ) gliomas, which can have distinct characteristics, disease outcome, and treatments compared to IDH wild-type ( IDH WT ) glioblastoma. We retrospectively examined Bev usage and efficacy at recurrence in a large single institution IDH MUT glioma cohort. Methods: We identified 138 consecutive IDH MUT glioma patients treated with Bev at recurrence at UCLA between 2005 and 2024. We stratified patients by 2016 WHO tumor diagnosis criteria. We examined how the timing of Bev initiation and other clinical variables affect Bev PFS and Bev OS. Bev PFS as determined by the treating clinician will be compared with our application prototype, Automated Imaging Response Evaluation System (AIRES), which is adapted to utilize RANO criteria. In addition, AIRES will be used to determine treatment response to be correlated with Bev PFS and Bev OS. Results: Time to Bev, which was defined as the time from initial surgery to Bev initiation, was shortest for AA and Grade 4 astrocytoma (G4 Astro). The median Bev PFS and Bev OS were 3.9 and 10.5 months, respectively. Bev PFS were 5.2, 4.6, and 3.3 months for the 1 st , 2 nd , and ≥3 rd recurrence groups, and Bev OS were 15.7, 13.5, and 8.3 months. Median Bev PFS, as determined by the treating clinician, for LO, LA, AO, AA, and G4 Astro were 2.8, 3.7, 6.2, 3.9, and 3.6 months, respectively. The median Bev OS were 7.0, 7.8, 9.5, 12.6, and 10.5 months. Among the 138 patients, the vast majority of patients initiated Bev treatment due to have contrast-enhancing tumor. 98 (71%) were first treated with Bev as a combined therapy, while 40 (29%) were treated with Bev monotherapy. 11 patients were concurrently treated with IDH inhibitors and Bev. The average daily corticosteroid (dexamethasone) dose reduction after 2 months of Bev treatment was 2.6 mg. Current results and Bev PFS are based on clinical criteria, while AIRES determined Bev PFS and response analysis are underway. Conclusions: Unlike our prior study on predominantly IDH WT GBM, later initiation of Bev was associated with diminished efficacy. Amongst the various diagnoses at Bev initiation, we found similar Bev PFS and Bev OS, except unexpectedly, we found AO to have improved Bev PFS and Bev OS compared to LO. Comparison with AIRES to determine Bev PFS and response according to RANO criteria is ongoing to finetune and validate results. Corticosteroid dose reduction was common but not associated with Bev PFS and Bev OS. With the recent increased use of IDH inhibitor treatment, this cohort, as a mostly IDH inhibitor naïve group, may prove valuable as a comparator arm to determine whether IDH inhibitor treatment affects Bev efficacy.