In response to major changes in diagnostic algorithms and the publication of mature results from various large clinical trials, the European Association of Neuro-Oncology (EANO) recognized the need to provide updated guidelines for the diagnosis and management of adult patients with diffuse gliomas. Through these evidence-based guidelines, a task force of EANO provides recommendations for the diagnosis, treatment and follow-up of adult patients with diffuse gliomas. The diagnostic component is based on the 2016 update of the WHO Classification of Tumors of the Central Nervous System and the subsequent recommendations of the Consortium to Inform Molecular and Practical Approaches to CNS Tumour Taxonomy — Not Officially WHO (cIMPACT-NOW). With regard to therapy, we formulated recommendations based on the results from the latest practice-changing clinical trials and also provide guidance for neuropathological and neuroradiological assessment. In these guidelines, we define the role of the major treatment modalities of surgery, radiotherapy and systemic pharmacotherapy, covering current advances and cognizant that unnecessary interventions and expenses should be avoided. This document is intended to be a source of reference for professionals involved in the management of adult patients with diffuse gliomas, for patients and caregivers, and for health-care providers.

Rindopepimut (also known as CDX-110), a vaccine targeting the EGFR deletion mutation EGFRvIII, consists of an EGFRvIII-specific peptide conjugated to keyhole limpet haemocyanin. In the ACT IV study, we aimed to assess whether or not the addition of rindopepimut to standard chemotherapy is able to improve survival in patients with EGFRvIII-positive glioblastoma. In this randomised, double-blind, phase 3 trial, we recruited patients aged 18 years and older with glioblastoma from 165 hospitals in 22 countries. Eligible patients had newly diagnosed glioblastoma confirmed to express EGFRvIII by central analysis, and had undergone maximal surgical resection and completion of standard chemoradiation without progression. Patients were stratified by European Organisation for Research and Treatment of Cancer recursive partitioning analysis class, MGMT promoter methylation, and geographical region, and randomly assigned (1:1) with a prespecified randomisation sequence (block size of four) to receive rindopepimut (500 μg admixed with 150 μg GM-CSF) or control (100 μg keyhole limpet haemocyanin) via monthly intradermal injection until progression or intolerance, concurrent with standard oral temozolomide (150-200 mg/m2 for 5 of 28 days) for 6-12 cycles or longer. Patients, investigators, and the trial funder were masked to treatment allocation. The primary endpoint was overall survival in patients with minimal residual disease (MRD; enhancing tumour <2 cm2 post-chemoradiation by central review), analysed by modified intention to treat. This trial is registered with ClinicalTrials.gov, number NCT01480479. Between April 12, 2012, and Dec 15, 2014, 745 patients were enrolled (405 with MRD, 338 with significant residual disease [SRD], and two unevaluable) and randomly assigned to rindopepimut and temozolomide (n=371) or control and temozolomide (n=374). The study was terminated for futility after a preplanned interim analysis. At final analysis, there was no significant difference in overall survival for patients with MRD: median overall survival was 20·1 months (95% CI 18·5-22·1) in the rindopepimut group versus 20·0 months (18·1-21·9) in the control group (HR 1·01, 95% CI 0·79-1·30; p=0·93). The most common grade 3-4 adverse events for all 369 treated patients in the rindopepimut group versus 372 treated patients in the control group were: thrombocytopenia (32 [9%] vs 23 [6%]), fatigue (six [2%] vs 19 [5%]), brain oedema (eight [2%] vs 11 [3%]), seizure (nine [2%] vs eight [2%]), and headache (six [2%] vs ten [3%]). Serious adverse events included seizure (18 [5%] vs 22 [6%]) and brain oedema (seven [2%] vs 12 [3%]). 16 deaths in the study were caused by adverse events (nine [4%] in the rindopepimut group and seven [3%] in the control group), of which one-a pulmonary embolism in a 64-year-old male patient after 11 months of treatment-was assessed as potentially related to rindopepimut. Rindopepimut did not increase survival in patients with newly diagnosed glioblastoma. Combination approaches potentially including rindopepimut might be required to show efficacy of immunotherapy in glioblastoma. Celldex Therapeutics, Inc.

The European Association for Neuro-Oncology guideline provides recommendations for the clinical care of adult patients with astrocytic and oligodendroglial gliomas, including glioblastomas. The guideline is based on the 2016 WHO classification of tumours of the central nervous system and on scientific developments since the 2014 guideline. The recommendations focus on pathological and radiological diagnostics, and the main treatment modalities of surgery, radiotherapy, and pharmacotherapy. In this guideline we have also integrated the results from contemporary clinical trials that have changed clinical practice. The guideline aims to provide guidance for diagnostic and management decisions, while limiting unnecessary treatments and costs. The recommendations are a resource for professionals involved in the management of patients with glioma, for patients and caregivers, and for health-care providers in Europe. The implementation of this guideline requires multidisciplinary structures of care, and defined processes of diagnosis and treatment.

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.