Probability Maps of Glioblastoma Indicate Variation in Surgical Decisions Between 10 Surgical Teams

INTRODUCTION The aim of glioblastoma surgery is to maximize the extent of resection, while preserving functional integrity. Standards are lacking for surgical decision-making and consequently surgical strategies may differ between neurosurgical teams. In this study, we quantitated and compared surgical decision-making throughout the brain between neurosurgical teams for patients with a glioblastoma using probability maps. METHODS All adults with first-time glioblastoma surgery in 2012-2013 from 10 tertiary referral centers for neurooncological care were included in this study. For each patient, pre- and postoperative tumor were manually segmented on MRI and aligned to standard brain space. Resection probability maps and biopsy probability maps were constructed in 1 mm resolution for each team's cohort. Brain regions with differential biopsy and resection results between teams were identified. RESULTS The study cohort consisted of 931 patients of whom 293 received a biopsy and 638 a resection. Biopsy probability maps demonstrated differences between teams in biopsy rate per brain location, such as for the left precuneus and superior parietal lobule, indicating variation in biopsy decisions. Resection probability maps demonstrated differences between teams in residual tumor rate per brain location, such as for the left saggital striatum and neighboring posterior corpus callosum, indicating variation in resection decisions. CONCLUSION Biopsy and resection probability maps indicate treatment variation between teams for patients with a glioblastoma. This conveys useful objective arguments for quality of care discussions between surgical teams for these patients