Fig. 6

Adult diffuse glioma molecular subtype predictions for the REMBRANDT study. A Predictions for all REMBRANDT patients show typical median survival for IDH-wildtype glioblastoma but worse outcomes for predicted IDH-mutant astrocytomas and oligodendrogliomas compared to those in the TCGA. B Unlike histologically-defined glioblastomas, histological astrocytomas and oligodendrogliomas in the REMBRANDT fair significantly worse than TCGA histological astrocytomas and oligodendrogliomas. C An 85% threshold for chromosomes 1p, 1q, 19p, and 19q successfully captures the densest area of patients who pass a 50% threshold for 1p/19q-codeletion. D Within the cohort of REMBRANDT histological astrocytomas, predicted IDH-wildtype tumors fair significantly worse than predicted IDH-mutant astrocytomas. E Within the cohort of REMBRANDT histological glioblastomas, predicted IDH-wildtype tumors’ outcomes are significantly worse than predicted IDH-mutant astrocytomas. F Predicted IDH-wildtype tumors do not show a subpopulation of young patients that would indicate the presence of pediatric-type tumors or misclassified IDH-mutant astrocytomas. G SCNA profiles of predicted IDH-wildtype tumors that live longer than 3 years. Three tumors show + 7/ − 10, of which two show co-gain of chromosomes 19 and 20, and are likely IDH-wildtype tumors. The remaining tumor shows mild losses on chromosome 10. H SCNA profiles of predicted IDH-mutant astrocytoma that live less than 12 months. None show SCNA characteristics of IDH-wildtype glioblastoma, suggesting that they are correctly classified IDH-mutant astrocytomas or possibly IDH-wildtype glioblastomas with TERTp mutations