Genomic characterization of patient-derived xenografts of glioblastoma
Glioblastoma Multiforme (GBM) is an aggressive brain tumor with a median survival rate of 14.6 months. However, some patients do not respond well to standard care, resulting in a decreased survival rate from the already poor prognosis. GBM treatment is especially difficult due to the heterogeneity and invasiveness of the tumors. Patient-derived tumor xenografts (PDX), cancerous tissue from a patient that is engrafted into immune-compromised rodents, can be used to facilitate the development of novel drug therapies and predict patient tumor response to anti-cancer agents. The goals of this project are to accurately identify variants from next generation sequencing data, to perform data mining to unveil the most frequent genetic alterations, and finally to present the findings through a web portal. First, exome and RNA sequencing were performed for 90 patient xenografts. To ensure accurate downstream analysis, mouse reads were removed (in silico). Single nucleotide variants (SNV), copy number alterations (CNA), and potential gene fusions were then called from the data, which was contained in variant calling format (VCF) files. Processing and analysis of the VCF files revealed a variety of mutations that were common among PDX samples. The most frequent SNVs among all 90 samples occurred on several genes with relatively high frequency. For instance, SNVs on the POU6F2 gene, which encodes a tumor suppressor protein, and the KAZN gene, which encodes a protein that plays a role in various functions such as desmosome assembly and cytoskeletal organization, were found to occur in approximately half of all GBM PDX samples. Ultimately, genomic characterization of PDX will enable researchers to identify potential biomarkers and test novel therapies that are tailored to a patient’s unique genome.