P01.007 Absence of Correlation between Necrosis Volume and Overall Survival in Glioblastoma

Abstract Background Necrosis is an important feature of glioblastoma. It is commonly accepted that it reflects the malignancy of the tumor and correlates with survival. We tested this assumption. Material and Methods 140 patients were included in a retrospective study of histologically confirmed glioblastoma patients. All patients were treated according to the “Stupp” protocol, after biopsy or resection. Volumetric analysis was performed using semi-automated segmentation with the S7 StealthStation software (Medtronic) on MPRAGE (Magnetization Prepared-Rapid Echo Gradient) images with 0.9mm thickness (1.5T). MGMT-promotor-gene methylation status was determined. Statistical analyses were performed with SAS software (version 9.4, basic edition, 2012–2016, SAS Institute Inc., Cary, NC, USA). Two-sided Mann-Whitney U test was used to compare mean ranks of tumor necrosis volume between MGMT groups. Separate Cox proportional hazard models were fitted, each adjusted for the type of procedure, to assess the lineair relationship of tumor necrosis volume with mortality hazard. Results Mean overall survival was 19,3 months. In 107 patients, volumetric analysis could be performed. There was no difference in mean necrosis volume between the MGMT-methylated group and non-methylated group. There was no difference in OS between the group with (n= 93) and without necrosis (p = 0.195). Necrosis volume at diagnosis was not correlated with mortality hazard (p = 0.484) nor when adjusted for surgical procedure (p = 0.473). Conclusion In this retrospective study using semi-automated volumetric analysis, there is no correlation between necrosis volume at diagnosis of glioblastoma and mortality hazard.