DESCRIPTION (provided by applicant): Neuroblastoma is a common and lethal pediatric malignancy, but the genetic events that initiate tumorigenesis are largely unknown. We hypothesize that neuroblastoma is a complex disease that results from the interaction of mutant alleles with relatively low to moderate effect on tumor initiation. We now plan to discover neuroblastoma susceptibility genes by performing a definitive whole genome association study in ~5000 neuroblastoma patients from the Children's Oncology Group using a high-density single nucleotide polymorphism (SNP)-based replicative study design. We will compare our cases to two control sets: both pediatric non-cancer patients accessioned from our institution, and adult control subjects matched by region across the country. We propose four Specific Aims: First, we will perform a whole genome scan for association of neuroblastoma with SNPs and SNP haplotypes in 2000 neuroblastoma cases and an equal number of matched controls (both children and adults in parallel analyses). Over 550K SNPs will be surveyed, and we anticipate identifying 25,000-50,000 candidate SNPs for further evaluation. Second, we will identify true disease-associated SNP alleles using a customized genotyping platform enriched for haplotype analyses in an independent set of 2000 cases and again two large control sets. By leveraging the HapMap project and controlling for population substructure, we anticipate that this Aim will identify 10-20 genomic regions as candidates for association with neuroblastoma. Third, we will validate at least 5-10 disease-associated regions in a final independent sample set of 675 cases and two sets of ~675 controls. We will again use a gene-centric haplotyping approach to determine true disease associated variants and candidate genes. Finally, we will definitively identify neuroblastoma predisposition genes through direct resequencing of candidate regions in a carefully selected set of 100 neuroblastoma cases and 100 controls, each equally proportioned between the presence or absence of the SNP/haplotype variant associated with neuroblastoma. Our large panel of tumor reagents and genomics databases will be leveraged to assist in region prioritization and identification of candidate genes for further analysis. The successful completion of this project will provide insight into the underlying genetic etiology of neuroblastoma tumorigenesis. We ultimately plan to translate the discovered neuroblastoma predisposition genes into a prognostic biomarkers and/or a target for new treatment approaches. In addition, the data generated here will rapidly be made available to any academically qualified petitioner interested in associating the SNP genotypes with the robust phenotypic information that we have captured including clinical characteristics, tumor biology, response to therapy and disease outcome. Finally, this project should also help catalyze the field of childhood cancer applied genomics research, and if past lessons from other childhood cancers are repeated, it will identify genes that are fundamentally important in human cancer in general. PUBLIC HEALTH RELEVANCE: Neuroblastoma is a common and lethal childhood cancer for which the genetic basis is poorly understood. Our genome-wide association study is designed to discover the most common variations in the human genome that lead to the development of neuroblastoma. These insights will lead to new molecular diagnostic assays and/or new treatments for this frequently devastating malignancy of young children.