Abstract SY25-03: Host and tumor genetic architecture influence clinically relevant oncogenic pathways in neuroblastoma

Abstract

Abstract Cancer research is at a critical crossroads as genomic technologies have enabled potentially paradigm-shifting discoveries in host genetic alterations that lead to cancer, as well as the major oncogenic drivers that somatically arise and shape the ultimate tumor phenotype. Approaches such as genome-wide association studies (GWAS) to discover heritable susceptibility variants have largely occurred in parallel with tumor profiling efforts designed to discover somatically acquired mutations. We see tremendous opportunity in integrating these efforts under the broad hypothesis that the host's genetic blueprint not only influences susceptibility to cancer, but also significantly influences the cancer phenotype at diagnosis and the pathways upon which the cancer is dependent. By focusing our efforts on the pediatric cancer neuroblastoma, we are directly addressing an important childhood malignancy that continues to exact disproportionate morbidity and mortality. However, we have also shown that this disease also serves as a remarkable model for the study of human cancers in general, and we think that lessons learned here are likely to be more broadly applicable as we are addressing the concept that somatic mutations alone will not define the entire spectrum of oncogenic vulnerabilities in cancer cells. Until recently, the etiology of neuroblastoma was unknown. By studying the 1% of neuroblastoma cases that occur within families, we have discovered ALK and PHOX2B as the two major hereditary predisposition genes.1,2 In parallel, by studying the 99% of cases that occur sporadically in our ongoing GWAS, for which we have now analyzed 4,000 cases and 10,000 controls, we have shown that normal genetic variation at multiple loci influences neuroblastoma susceptibility.3-7 We have several other recently validated associations that are not yet published, including replicated signals in LIN28B and HACE1. Significant lessons from the GWAS to date include: 1) Neuroblastoma is a genetic disease; 2) every signal identified to date is intragenic; 3) associations are enriched in phenotypic subsets; and 4) many of the genes identified by the neuroblastoma GWAS not only influence malignant transformation of a developing neuroblastic cell, but also contribute to the maintenance of the malignant phenotype. For example, SNP variations in the LMO1 gene are highly associated with overexpression of this transcriptional co-factor, and genetic manipulation of expression shows that it functions as an oncogenic driver in a genotype specific manner.7 In other words, our data suggest that subjects harboring LMO1 risk alleles are much more likely to have tumors dependent on LMO1-influenced transcriptional regulation suggesting that oncogene addiction is also hardwired in the heritable genome. In parallel, we and others have shown that somatic activating mutations in the ALK kinase domain occur in ∼10% of sporadic cases.2,8-10 To seek other potential somatically acquired mutations that may be therapeutically relevant in the neuroblastoma coding genome, we have resequenced 99 high-risk neuroblastoma cases using a combination of whole exome, genome and transcriptome sequencing. Tumors contained a median 0.56 non-silent mutations per megabase of coding DNA, one of the lowest rates reported in cancer to date. ALK was the most commonly mutated gene (9%), and no other gene showed mutation frequency rates >5%. These data clearly indicate that somatic mutation in neuroblastoma is relatively rare, leading to the concept that other mechanisms, such as host polymorphic variation or germline mutations in key regulatory regions, play an important role in tumorigenesis. Ongoing work to be discussed at the symposium include efforts to identify and validate rare variants in the host genome that impact neuroblastoma susceptibility, understanding the functional relevance of DNA variation in the developing neuroblastoma and in tumor tissues, completion of our cataloguing of somatic mutations in high-risk disease, and efforts to understand the impact of non-coding DNA alterations in the host in tumor genomes on neuroblastoma clinical phenotype. Efforts to translate these efforts into deliverables for the clinic will be emphasized.