Genomic and Molecular Characterization of Brainstem Glioma

Abstract

<p>Gliomas of the brainstem and closely surrounding structures are a devastating form of brain cancers with some of the highest mortality rates among all cancers. The delicate structures surrounding these tumors make surgical resection rare due to the risk of perioperative mortality or paralysis. A lack of available tissue has created a dearth of genomic information regarding gliomas effecting specific structures in the brainstem impacting options for intervention.</p><p>This work comprises two major projects of brainstem gliomas. The first part is to utilize next generation sequencing technique to discover novel oncogenes, and to use epigenetic features to characterize brainstem gliomas. The second part is to utilize integrated analysis of whole genome sequencing, methylation microarray, RNA sequencing data to establish the classification of brainstem gliomas.</p><p>In the first part we completed exomic and Sanger sequencing on 33 brainstem gliomas and 17 thalamic gliomas to reveal major findings regarding their respective oncogenomes. We reported the presence of IDH1 mutations occurring in brain- stem gliomas exclusive of H3.3 mutations. We further revealed somatic mutations to PPM1D, the gene encoding for the Protein phosphatase 1D in 37.5% of brainstem gliomas harboring H3.3 mutations. We determined that PPM1D mutations are mu- tually exclusive with TP53 mutations in brainstem gliomas. This was the largest body of genomic work completed on specific loci within the brainstem providing new insights into their respective oncogenomes and presents PPM1D as a new oncotarget.</p><p>In the second part we then performed a comprehensive study incorporating com- bined epigenetic and genomic analyses with clinical features, aiming to understand the genetics and molecular biology of this disease. From DNA methylation data, we identified four distinct clusters termed H3-Pons, H3-Medulla, IDH, and PA-like, each associated with unique genomic and clinical profiles. The majority of tumors within H3-Pons and H3-Medulla harbored H3F3A mutations, but showed distinct methy- lation patterns that correlated with anatomical localization of these tumors within the pons or medulla, respectively. Clinical data also showed significantly different overall survival between these clusters, and pathway analysis demonstrated different mechanisms in these samples. In summary, we performed a comprehensive molecular profiling study on a large brainstem glioma cohort, including Diffuse Intrinsic Pon- tine Gliomas. Our findings indicate that the integration of DNA methylation data may facilitate a better understanding of brainstem glioma classification and guide future studies for the development of novel treatments for this disease.</p>