Spatio-temporal evolution of the primary glioblastoma genome (aCGH)

Tumor recurrence following a standard treatment is the major cause of mortality for glioblastoma (GBM) patients. However, insights on the evolutionary process of the tumor have been limited due to the lack of longitudinally sampled cases. Here, we describe our genomic analyses of 38 GBM patients with pre- and post-treatment samples for each individual (78 tumor samples in total; aCGH data were obtained for 36 among the 78). A substantial shift in the landscape of driver alterations was associated with distant appearances of the recurrent tumors from the initial tumor, suggesting that the genomic profile of an initial tumor can mislead targeted therapies for the distant recurrent tumor. In addition, in contrast to the previous work on IDH1-R132H low-grade gliomas, our GBM patients rarely developed hypermutation following the standard treatment, supporting the safety of temozolomide for IDH1-wild type, primary GBMs under the current standard regimen. aCGH experiments were performed for 36 pre- or post-treatment human glioblastoma tissues using the Agilent Human Whole Genome 180K CGH microarray according to manufacturer's protocol. Random human male normal tissue gDNA (Cy5 labeled) was used as a reference for each experiment.