Validation of MRI patterns of glioblastoma multiforme to predict MGMT methylation

Abstract

Background: Glioblastoma Multiforme (GBM) is the most frequent and deadliest of malignant brain tumors with a mean overall survival of 24% at first year. In the last decade magnetic resonance imaging (MRI) has become an essential tool for diagnosis, but is still not perfect as the final diagnostic is based on the histopathological analysis. In the last years numerous studies have observed that radiological characteristics may reflect the biological features of GBM and may be associated with genetic aberrations and molecular alterations that occur in tumorigenesis. Radiological patters have been also related to prognosis and patient outcomes. In that way, is logical to think that there is a link between some radiological patterns, molecular properties and, because of that, prognosis. MRI could complement the use of biopsy as a diagnostic instrument in the molecular diagnosis of glioblastoma, and the best way to predict patient’s prognosis and to give realistic expectation to the patients. Objective: To validate MRI patterns of Glioblastoma Multiforme to predict the IDH and MGMT status. Design: A multicentric, observational, prospective cohort with 3 years of follow-up. Three third level hospitals of Institut Català d’Oncologia (ICO): Duran i Renyals hospital, Germans Trias i Pujol hospital, and Dr. Josep Trueta hospital will participate in this project. Participants: people older than 18 years old that have been diagnosed of GBM in the 2 years of duration of the sample collection, and had no documented history of previous brain tumors. Methods: Information of 240 patients with confirmed GBM will be analysed, including pre and postoperative MRI and the clinical features. The data will be obtained mainly by the informatic date base, with the patient’s medical histories, that links the hospitals. For the statistical analysis we will use a chi-squared test, a T-student or Mann-Withney’s U depending on the symmetry of the variables. The progression-free survival (PFS) and overall survival (OS) will be summarized with the Kapplan-Meier survival curve estimator and then asses it with the Log-rank test. A multivariate analysis will be also done adjusting using logistic regressions and the Cox regression