Abstract
Glioblastoma is an aggressive tumor that is clinically and pathologically heterogeneous, and as such, remains challenging to treat. Response to initial standard treatment is widely variable and patient specific. We describe a cohort of 11 newly diagnosed patients enrolled in the 3D-PREDICT study at a single institution who underwent maximal safe surgical resection with tumor tissue collected prospectively for ex vivo cell culture assays against a panel of commonly used agents, including temozolomide. Nine of eleven patients received concurrent radiation + temozolomide followed by adjuvant temozolomide as per current standard of care. Two patients progressed immediately following concurrent chemoradiation; pseudoprogression was ruled out with short interval repeat imaging and clinical deterioration with pathology confirming recurrent tumor on re-resection in one patient. Median follow-up was 10 months (range 2-18). Outcomes were assessed retrospectively using Kaplan-Meier time-to-event curves, separating into two groups the temozolomide responders (n=4) and temozolomide nonresponders (n=7), as defined by the previously validated assay. The event was reached when radiographic tumor recurrence/progression occurred. Due to limited sample size, statistical significance was not reached, but a trend toward longer time to recurrence was noted among the temozolomide responder group. Continued experience with this tool may help clinicians predict which patients with newly diagnosed glioblastoma will respond well to initial treatment with temozolomide and those that may be more appropriate for clinical trial enrollment, independent of MGMT promoter methylation status. Longer term studies with a larger number of patients will help to determine the true significance of this drug response prediction assay.
AT-20 * SINGLE INSTITUTION RETROSPECTIVE REVIEW OF PERFORMANCE STATUS AND CORTICOSTEROID USE IN NEWLY DIAGNOSED GLIOBLASTOMA (GBM) PATIENTS (PTS) TREATED ON BEVACIZUMAB (BEV)