SURG-15. INTEGRATING MOLECULAR MARKERS AND EXTENT OF RESECTION FOR RISK STRATIFICATION OF PATIENTS WITH NEWLY-DIAGNOSED GLIOBLASTOMA: A MULTICENTRE STUDY

Abstract

BACKGROUND: Multiple studies have shown that extent of surgical resection (EoR) is an independent prognostic factor for patients with newly-diagnosed glioblastoma. Previous work has proposed the inclusion of EoR in a risk stratification algorithm but does not incorporate recent advances in the molecular characterization of tumors. OBJECTIVE: To develop an integrative risk stratification scheme that incorporates clinically relevant molecular data and EoR with classic prognostic variables to individualize prognosis and guide treatment and research. METHODS: We reviewed all consecutive cases of confirmed newlydiagnosed glioblastoma who were operated upon between January 1, 2012 and December 31, 2014 at two tertiary academic centres. Variables including age, sex, KPS, tumour location, presenting symptoms, treatment history, dates of progression and reoperation, as well as MGMT promoter methylation (MGMT-M), IDH, 1p/19q codeletion, and ATRX status were recorded. Computer-assisted volumetric analysis of pre- and post-operative MRIs allowed calculation of pre-operative tumour burden, residual disease, and %EoR. RESULTS: Preliminary results from review of 63 out of 297 cases diagnosed during the study period showed patients with EoR ≥ 95% and positive MGMT-M had the longest median overall survival (23.7 months), but the benefit of MGMT-M was not seen at lower EoR. CONCLUSION: The combination of MGMTM and EoR ≥ 95% is synergistic in improving patient survival, possibly reflecting the improved efficacy of chemotherapy at lower residual tumour volumes. Review of remaining cases and recursive partitioning analysis(RPA) are pending, and will allow development of an integrative risk stratification algorithm.