No abstract

Introduction: Bevacizumab has demonstrated activity in glioblastoma (GBM), but the true benefits and optimal dose-schedule are debated. A lower dose-schedule than standard-dose bevacizumab (10 mg/kg 2-weekly) might offer similar benefits with lower costs. At our Institution, patients are randomly assigned at time of primary diagnosis to Neuro-Oncologists, who have varying practices in terms of bevacizumab dose-schedule upon progression. Methods: In a retrospective analysis we examined overall survival (OS), measured from first administered bevacizumab dose until death, according to dose-schedule.Patients with de novo WHO Grade IV GBM who received standard-or reduced-dose (5 mg/kg 2-weekly) bevacizumab were included. MGMT methylation status and time from diagnosis to bevacizumab start were examined as prognostic variables.Clinical benefit and a comparative cost analysis were assessed. Results: In total, 1127 bevacizumab doses were administered to 118 patients [Median: 7, Range: 1-44]. Median OS (mOS) was 5.8 months. 69 (59%) patients received standard-dose bevacizumab (mOS: 5.97 months) and 49 patients received reduced-dose (mOS: 5.7 months). No statistically significant difference in OS between dosing schedule was seen (HR: 1.11, P-value: .584). Patients with MGMT methylated tumors (43%) had improved OS compared to those with unmethylated tumors; 7.03 vs 4.97 months (HR: 0.61, P-value: .027). If all patients were treated with reduced-dose bevacizumab, an estimated €2.4M cost reduction would be observed. Conclusions: In this retrospective study, reduced-dose bevacizumab schedule resulted in similar OS to standard-dose bevacizumab monotherapy with substantial cost savings. MGMT methylation appears to convey a survival benefit in the setting of bevacizumab treatment for progressive GBM. K E Y W O R D Sbevacizumab, cost analysis, glioblastoma, MGMT, overall survival, reduced dose 470 | GLEESON Et aL.

Gastrointestinal (GI) malignancies are a major global health burden, with high mortality rates. The identification of novel therapeutic strategies is crucial to improve treatment and survival of patients. The poly (ADP-ribose) polymerase (PARP) enzymes involved in the DNA damage response (DDR) play major roles in the development, progression and treatment response of cancer, with PARP inhibitors (PARPi) currently used in the clinic for breast, ovarian, fallopian, primary peritoneal, pancreatic and prostate cancers with deficiencies in homologous recombination (HR) DNA repair. This article examines the current evidence for the role of the DDR PARP enzymes (PARP1, 2, 3 and 4) in the development, progression and treatment response of GI cancers. Furthermore, we discuss the role of HR status as a predictive biomarker of PARPi efficacy in GI cancer patients and examine the pre-clinical and clinical evidence for PARPi and cytotoxic therapy combination strategies in GI cancer. We also include an analysis of the genomic and transcriptomic landscape of the DDR PARP genes and key HR genes (BRCA1, BRCA2, ATM, RAD51, MRE11, PALB2) in GI patient tumours (n = 1744) using publicly available datasets to identify patients that may benefit from PARPi therapeutic approaches.

e13515 Background: Bevacizumab (BEV) has demonstrated activity in GBM, but the benefits have not been clearly defined in prospective randomised phase III trials. Some patients continue BEV for extended periods but little is known about predictors of clinical benefit. It has been suggested that a less intensive dose schedule might offer similar benefits. Methods: In a retrospective analysis of a National Neuro-Oncology centre database we examined biomarkers of BEV benefit on overall survival (OS), including age at diagnosis, time from diagnosis to BEV start, MGMT methylation and IDH-1 status. We compared those who derived clinical benefit (OS of ≥6-months) to those with limited clinical benefit (OS <6 months). At our Institution, there is variable practice between Neuro-Oncologists in terms of BEV dose schedule. In this homogeneous patient population, we examined whether OS was similar for patients who received standard BEV dosing (10mg/kg q 2wks or 15mg/kg q 3wks) or reduced intensity BEV (5mg/kg q 2wks or 7.5mg/kg q 3wks). Results: Between January 1st 2010 and Dec 31st 2016, 170 patients received BEV for progressive GBM. Median OS was 5.5 months (Range: 0.5-54). OS was 42%, 18% and 2% at 6-, 12- and 24-months, respectively. No significant predictors of clinical benefit were identified (table). Importantly, median OS was similar for patients who received standard (N=94) vs. reduced intensity (N=76) BEV (5.2 vs. 5.6mo respectively), p=0.74. Conclusions: In this large cohort of patients, we did not identify predictors of benefit from BEV. OS was similar in patients who received standard or reduced intensity BEV. Given the cost of BEV, these results have important implications for value in cancer care. [Table: see text]