SURG-11THE IMPACT OF PRE-OPERATIVE TUMOR FEATURES ON RESECTION AND SURVIVAL OUTCOMES IN GLIOBLASTOMA: A PHASE II FLUORESCENCE-GUIDED SURGERY STUDY

Abstract

BACKGROUND: The survival benefit provided by “more complete” resection of glioblastoma (GBM) remains under debate. Part of this debate is likely due to the lack of reliable tools for pre- and post-surgical volumetric tumor measurement. In this study, we used a validated, semi-automated tumor segmentation tool to evaluate the relationship between pre-operative tumor morphology metrics, resection outcomes, and survival outcomes in a 5-aminolevulinic acid (5-ALA) fluorescence-guided surgery (FGS) trial. METHODS: Thirty-one patients from a prospective Phase II trial evaluating 5-ALAFGSfor newly diagnosed GBM were assessed. Tumors were segmented pre-operatively using a semi-automated method to assess morphological features, such as surface area (SA), SAto volume ratio (SAVR), and percent necrosis (NV%), as well as post-operatively to evaluate extent-of-resection (EOR) and residual tumor volume (RTV). Multivariate analysis was performed using a general linear model to assess the effects of multiple covariates on RTV and to describe resectability in terms of pre-operative tumor features. The impact of morphological, clinical, and resection measures on progression free survival (PFS) and overall survival (OS) were evaluated using univariate and multivariate Cox proportional hazard regression. RESULTS: Median EOR and RTV were 94.3% and 0.821 cm3, respectively. Pre-operative tumor volume, SA, SAVR, and NV% exhibited statistically-significant associations with RTV. In addition to age and O-6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status, preoperative tumor SA, SAVR, NV%, and volume were shown to impact the likelihood of complete resection. Pre-operative SAVR and RTV were significantly associated with OS, even controlling for the known confounders of age, KPS, and MGMT methylation. CONCLUSIONS: This study supports claims that 5-ALA FGS is more effective than conventional microsurgery in decreasing tumor burden in GBM. Moreover, morphological indices describing tumor shape and composition may impact not only resection outcomes, but also patient survival, and should be considered in tumor resection studies.