OS08.7.A NEUROSURGEONS’ VIEWS ON SURGICAL TRIALS FOR GLIOBLASTOMA: ATTITUDES, PRACTICES, AND WILLINGNESS TO RANDOMIZE

Abstract BACKGROUND The scarcity of randomized controlled trials (RCTs) in surgical neuro-oncology tends to limit the level of evidence for clinical practice. Specifically in the surgical treatment of glioblastoma, controversies on the optimal operative treatment remain. To evaluate the degree of equipoise and feasibility of surgical RCTs for Glioblastoma (GBM), we conducted an international survey of neurosurgeons, exploring neurosurgeons’ habits, opinions on current quality of evidence, and willingness to randomize patients with GBM into a surgical trial under specific circumstances. MATERIALS AND METHODS We conducted a global survey among neurosurgeons. The survey was distributed through EORTC, EANS and CNS as well as several national neurosurgical societies. Chi-square test was used to assess the association between categorical variables. RESULTS In total, 328 neurosurgeons from 38 countries and 5 continents responded. Most participants (83.2%) worked in academic centers. 81.7% were male. All career stages (defined as years since completing training) were well represented in the survey, as were individual volumes of brain tumors operated yearly in current practice. Neuronavigation (98.5%), intraoperative neuromonitoring (80.2%), fluorescence (80.6%) and intraoperative ultrasound (60.1%) are the most frequently used technological aids. A majority of neurosurgeons (60.4%) appear to hold supramarginal resection(SM) as the optimal goal in GBM surgery, while 37.8% maximally aim to achieve a complete resection of enhancing tumor (CRET). Significant differences in these views exist across countries (p<0.001), and female neurosurgeons appear to tend more towards SM than male neurosurgeons (71.5% vs 57.5%) (p=0.02). A large majority (90.2%) conveyed that evidence for CRET is “strong” to “extremely strong”, while for SM, this was the case in 63.1%. 72.6% would be willing to randomize between CRET and SM, while 49.4% would be willing to randomize between partial resection and biopsy in case CRET is not feasible. Although statistically significant differences in practice and impression of evidence exist across countries, willingness to randomize in the provided scenarios was largely similar across countries. A paradoxical finding was that surgeons who conveyed that the evidence is currently weak were also less willing to randomize. (p<0.001)Conclusion: Although there is a paradoxical relationship between individual impression of strength of evidence and willingness to randomize, a majority of neurosurgeons practicing neuro-oncology worldwide appear to be willing to participate in randomized clinical trials addressing controversies in the surgical treatment of glioblastoma. Those results may inform the further design of randomized clinical trials in surgical neuro-oncology.