Patterns of Malignant Glioma Relapse after Discontinuing Bevacizumab Prior to Disease Recurrence (P03.138)

Abstract

Objective: To investigate whether discontinuing bevacizumab prior to tumor recurrence leads to radiogrpahic rebound, results in an invasive/multifocal recurrence, or leads to early tumor progression/poor survival. Background Bevacizumab (BEV) is an anti-angiogenic therapy approved for recurrent glioblastoma. In BEV responders, it is uncertain whether: 1.) the radiographic response represents anti-tumor effect or restoration of the blood-brain barrier; 2.) continued therapy causes a change to an invasive tumor phenotype; 3.) treatment can be stopped without compromising survival. Design/Methods: We retrospectively identified recurrent malignant glioma patients (>16 weeks after radiotherapy to minimize those with pseudoprogression) who achieved a radiographic response to BEV and discontinued BEV prior to disease recurrence. Results: 7 (5 glioblastoma, 2 anaplastic astrocytoma) were identified. Median age was 50 years. BEV was given at first (5), second (1), and fourth (1) recurrence. BEV was given in combination with cytotoxic chemotherapy (4) and as a single agent (3). Patients received a median of 7 (range 4-11) BEV treatments. BEV was discontinued for toxicity (4) and patient/physician choice (3). None received dexamethasone or other tumor treatment after BEV discontinuation. After stopping BEV, median time to recurrence and six-month progression free survival were 4.6 months (range 1.4-22.1) and 43%. respectively. The pattern of tumor recurrence was local (5), distal (1), and infiltrating/multifocal (0). One glioblastoma has not recurred 22.1 months after stopping treatment. Five were re-treated with BEV at recurrence and 4 had a radiographic response. 4/5 glioblastoma patients remain alive at a median follow-up of 9 months after initial discontinuation of BEV. Conclusions: A radiographic rebound effect was not observed. BEV re-treatment was associated with radiographic response despite prior exposure. Although treatment recommendations cannot be made based on this small, retrospective series, the encouraging survival of recurrent glioblastoma suggests that early discontinuation of BEV may be considered. Prospective studies are necessary to define optimal BEV schedule. Disclosure: Dr. Sherman has nothing to disclose. Dr. Raizer has received personal compensation for activities with Genentech, Merck, and Enzon as a speaker and advisory board member. Dr. Raizer has received research support from Genentech Inc., Novartis, Lilly, Myriad, Arno, and Med-Immune. Dr. Grimm has nothing to disclose.