Abstract 5253: Rare copy number variations (CNVs) influence neuroblastoma susceptibility

Abstract

Abstract Neuroblastoma is a pediatric malignancy of the developing sympathetic nervous system which is often lethal. Approximately 1% of cases are inherited in an autosomal dominant fashion, yet the remaining 99% appear to occur sporadically. We have previously demonstrated that common single nucleotide polymorphisms (SNPs) and common copy number variations (CNVs) at several loci are highly associated with neuroblastoma susceptibility (NEJM 2008, Nat Genet 2009, Nature 2009). By definition, common variants will each confer only a modest risk of developing a given disease. We hypothesized that rare variants/mutations play an important role in the etiology of sporadic neuroblastoma and have a strong influence on disease susceptibility. To test this hypothesis, we used data from our ongoing genome-wide association study (GWAS) to identify genes influenced by rare CNVs unique to neuroblastoma patients and test whether these genes are enriched in specific functional categories and/or pathways. We first identified a discovery set of 1,205 Caucasian neuroblastoma cases and 1,942 Caucasian healthy controls, each genotyped on the Illumina HumanHap550 or HumanHap610 SNP arrays. We observed large (> 1 Mb) CNVs in 4.1% of cases (median size = 1.3 Mb; range = 1.0 - 23.6 Mb), including two non-overlapping constitutional deletions greater than 5-Mb at 11q. Large CNVs (> 1 Mb) were significantly less common in controls (p = 3.5 × 10−4; 2.3% controls). We next mapped CNVs for each sample to known protein coding genes captured in the RefSeq database (NRefSeq ∼ 32,000). We identified 219 genes overlapping CNVs in multiple cases, but not detected in any controls; these CNVs were individually rare and occurred in < 1% of cases. To investigate the biological relevance of this gene set, we tested for enrichment in categories defined by Gene Ontology (GO), SwissProt keywords, Interpro protein domains, PIR (International Protein Sequence Database), and KEGG pathways. We observed significant enrichment in “pattern specification” (p = 4.3 × 10−10), “homeobox domain” (p = 4.7 × 10−7) and “developmental protein” (p = 1.3 × 10−6). Importantly, the set of genes with CNVs uniquely identified in controls (N = 529) was random and not enriched in any functional category. Finally, we replicated these findings in an independent set of 363 cases and 2,219 controls (p = 1.5 × 10−4 - 0.02). Several candidate neuroblastoma predisposition genes emerged from this study including homeobox genes HOXA10 and HOXA11 which have been implicated in other human cancers via hyper-methylation. We conclude that rare CNVs present in the constitutional DNA of neuroblastoma patients contribute to disease susceptibility. These CNVs can be very large and/or preferentially affect developmental genes, many of which have been implicated in cancer. Ongoing efforts are focused on validating the most common CNVs and determining whether they represent inherited genetic variation or de novo mutational events. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 5253.