In this project we will compare the genomes of 2,920 children with glioma to 2,920 cancer-free children in order to discover the underlying genetic causes of this deadly pediatric brain cancer. We will identify genetic risk factors that are common in the population and which increase many children's risk of glioma, as well as genetic risk factors unique to affected individuals, and then validate identified risk factors in three additional populations. This type of study has never been attempted before, and has the potential to reveal biologic pathways underlying pediatric brain cancer risk, thereby identifying targets for future prevention and therapy.
Brain tumors are the most common solid tumor and the second most common malignancy in children. Nearly 2,000 pediatric gliomas (PG) are diagnosed annually in U.S. children under the age of 15, only half of whom survive into adulthood. 80% of survivors experience life-threatening conditions related to treatment, including stroke and second malignancies. Despite clear evidence of a genetic component underlying PG risk, little is known about heritable factors affecting this deadly brain tumor. As previously demonstrated for adult glioma, identification of robust and validated genetic risk factors can lead to improved risk stratification and reveal the biologic pathways fundamental to the disease pathogenesis. Previous studies seeking to identify genetic risk factors for PG were limited by small sample size. This obstacle rendered studies inadequate for the identification of authentic genetic associations in high-throughput fashion. Furthermore, technological limitations have forced prior studies to focus on common genetic variation, which may be only one component of the genetic origins underlying PG risk. The hypothesis that both rare and common genetic variation contribute to PG risk, and risk of specific subtypes, will be formally tested in this proposal. To achieve this, a population-based case-control study, nested within the California Birth Cohort (CBC), has been developed. Genome-wide analysis of common and rare genetic variants will be conducted using existing archived neonatal bloodspots from 2,920 Californian children diagnosed with PG between 1988 and 2013, and 1:1 matched controls. First, DNA from 300 children with malignant astrocytoma and 100 controls will undergo whole-exome sequencing (WES) to identify rare variants contributing to disease risk (Minor Allele Frequency<1%). Genes displaying significant enrichment of rare variants in affected children compared to CBC and public control exomes will be validated by targeted sequencing in an additional 675 malignant astrocytoma case children and 875 control children from the CBC. Next, 20,000 promising low-frequency variants (MAF 1-5%) identified from the WES will be added as custom content to a genome-wide genotyping array, already containing 818,000 common variants. DNA samples from all 2,920 CBC case children and 2,920 CBC control children will undergo genome-wide genotyping to perform an empirically-enriched genome-wide association study (eeGWAS). The eeGWAS analysis can identify both low-frequency and common variants underlying PG risk, and is statistically powered for both pooled and subtype-stratified analyses. Approximately 1,500 variants identified by the eeGWAS will undergo attempted replication in 1,210 case and 1,850 control children from three collaborating institutions. By leveraging the unique and mature resources within the Genetic Diseases Branch of the California Department of Public Health, this registry-based approach will yield an unprecedentedly large sample size. The identification of both rare and common variants underlying PG risk can expose new knowledge leading to improved care of children, adolescents, and young adults facing this diagnosis.