Identification de sous-groupes de glioblastome sur base de critères cliniques

Glioblastoma has terrible outcomes, but the disease evolution varies a lot between patients. We used a sample of 239 patients (EORTC 26981) and a validation sample of 93 patients (CHU Liège) to class patients in groups with the same risk of recurrence. 4 distinct subgroups were created using recursive partitioning analysis. In the future, genetic profile will be compared between these groups to find markers of aggressiveness. Moreover, survival after recurrence is closely related to subgroups, and a surgical resection of the recurrence shows positive effect in each subgroup.