Richard G. Grundy

PURPOSE: To define copy number alterations and gene expression signatures underlying pediatric high-grade glioma (HGG). PATIENTS AND METHODS: We conducted a high-resolution analysis of genomic imbalances in 78 de novo pediatric HGGs, including seven diffuse intrinsic pontine gliomas, and 10 HGGs arising in children who received cranial irradiation for a previous cancer using single nucleotide polymorphism microarray analysis. Gene expression was analyzed with gene expression microarrays for 53 tumors. Results were compared with publicly available data from adult tumors. RESULTS: Significant differences in copy number alterations distinguish childhood and adult glioblastoma. PDGFRA was the predominant target of focal amplification in childhood HGG, including diffuse intrinsic pontine gliomas, and gene expression analyses supported an important role for deregulated PDGFRalpha signaling in pediatric HGG. No IDH1 hotspot mutations were found in pediatric tumors, highlighting molecular differences with adult secondary glioblastoma. Pediatric and adult glioblastomas were clearly distinguished by frequent gain of chromosome 1q (30% v 9%, respectively) and lower frequency of chromosome 7 gain (13% v 74%, respectively) and 10q loss (35% v 80%, respectively). PDGFRA amplification and 1q gain occurred at significantly higher frequency in irradiation-induced tumors, suggesting that these are initiating events in childhood gliomagenesis. A subset of pediatric HGGs showed minimal copy number changes. CONCLUSION: Integrated molecular profiling showed substantial differences in the molecular features underlying pediatric and adult HGG, indicating that findings in adult tumors cannot be simply extrapolated to younger patients. PDGFRalpha may be a useful target for pediatric HGG, including diffuse pontine gliomas.

Recent sequencing studies have extensively explored the somatic alterations present in the nuclear genomes of cancers. Although mitochondria control energy metabolism and apoptosis, the origins and impact of cancer-associated mutations in mtDNA are unclear. In this study, we analyzed somatic alterations in mtDNA from 1675 tumors. We identified 1907 somatic substitutions, which exhibited dramatic replicative strand bias, predominantly C > T and A > G on the mitochondrial heavy strand. This strand-asymmetric signature differs from those found in nuclear cancer genomes but matches the inferred germline process shaping primate mtDNA sequence content. A number of mtDNA mutations showed considerable heterogeneity across tumor types. Missense mutations were selectively neutral and often gradually drifted towards homoplasmy over time. In contrast, mutations resulting in protein truncation undergo negative selection and were almost exclusively heteroplasmic. Our findings indicate that the endogenous mutational mechanism has far greater impact than any other external mutagens in mitochondria and is fundamentally linked to mtDNA replication.

BACKGROUND: Gene expression profiling is being widely applied in cancer research to identify biomarkers for clinical endpoint prediction. Since RNA-seq provides a powerful tool for transcriptome-based applications beyond the limitations of microarrays, we sought to systematically evaluate the performance of RNA-seq-based and microarray-based classifiers in this MAQC-III/SEQC study for clinical endpoint prediction using neuroblastoma as a model. RESULTS: We generate gene expression profiles from 498 primary neuroblastomas using both RNA-seq and 44 k microarrays. Characterization of the neuroblastoma transcriptome by RNA-seq reveals that more than 48,000 genes and 200,000 transcripts are being expressed in this malignancy. We also find that RNA-seq provides much more detailed information on specific transcript expression patterns in clinico-genetic neuroblastoma subgroups than microarrays. To systematically compare the power of RNA-seq and microarray-based models in predicting clinical endpoints, we divide the cohort randomly into training and validation sets and develop 360 predictive models on six clinical endpoints of varying predictability. Evaluation of factors potentially affecting model performances reveals that prediction accuracies are most strongly influenced by the nature of the clinical endpoint, whereas technological platforms (RNA-seq vs. microarrays), RNA-seq data analysis pipelines, and feature levels (gene vs. transcript vs. exon-junction level) do not significantly affect performances of the models. CONCLUSIONS: We demonstrate that RNA-seq outperforms microarrays in determining the transcriptomic characteristics of cancer, while RNA-seq and microarray-based models perform similarly in clinical endpoint prediction. Our findings may be valuable to guide future studies on the development of gene expression-based predictive models and their implementation in clinical practice.