Abstract
INTRODUCTION: Supramarginal resection has gained traction as a strategy to improve outcomes in glioblastoma. Accumulating retrospective evidence suggests a survival benefit, yet the efficacy, safety, and feasibility of supramarginal resection remain to be tested in randomized controlled trials. METHODS: In an interim analysis of multicentre randomized controlled trial with a parallel group design, we compare supramarginal resection with standard of care (SOC) removal of suspected glioblastoma in patients amendable to supramarginal resection. Here, supramarginal was defined as resection with the intent of extending 1 cm beyond the MRI detected tumor core. As patients were included based upon suspicion, a modified intention to treat (mITT) analysis for efficacy was performed in patients with verified glioblastoma. All randomized patients, the intention to treat (ITT) population, were included in the safety analyses including neurological function, complications, and cognition (evaluated with Mini mental status exam [MMSE]). RESULTS: There were 38 patients in the ITT group whereas 31 were valid randomizations of glioblastoma and included in the mITT. In the mITT, the median age was 61y and 74% were male. Neurological function and complications were similar between groups, but for cognition a decline of 2 or more points in MMSE were seen in 8 % in SOC compared with 38 % in the supramarginal group. Hypermethylated MGMT promoter were seen in 29 % in both treatment arms. In the entire mITT cohort, median survival was 32 months (95% CI 24-40). The hazard ratio for death was 0.96 (95% CI 0.27-3.33) in supramarginal group compared with SOC. CONCLUSIONS: Randomization to supramarginal resection versus SOC is feasible. We provide preliminary data on the efficacy and safety of supramarginal resection. Based upon our preliminary results, a widespread approach of supramarginal resection outside clinical trials is not encouraged. These results should be considered when designing future clinical trials.