In the US, 54,640 head and neck cancer (HNC) cases and 11,520 deaths due to HNC occur each year. HNCs are malignancies that arise in the oral cavity, pharynx and larynx. Tobacco and alcohol are the two major risk factors for HNC, while human papillomavirus (HPV) infection is an important risk factor for oropharyngeal cancer. We conducted a HNC genome wide association study (GWAS), but the top hits were in the alcohol metabolism gene families, which we had already reported in our candidate gene projects. While further GWAS will identify additional common genetic variants with small effects, and undoubtedly some of the familial risk will be attributed to non-genetic factors such as tobacco and alcohol exposure, it is unlikely that these additional factors will increase the explained component of familial risk to over 50%. New approaches beyond GWAS, targeting either low frequency genetic variants (MAF=1.0-4.9%), or susceptibility genes harboring many rare variants (individual MAFs <1%), are necessary. It is our hypothesis that a significant portion of the unexplained familial risk of HNC is due to low frequency or rare genetic variants in multiple genes conferring relative risks ranging from approximately two-fold to greater than five-fold that cannot be detected by genome-wide SNP association approaches (even using several million SNPs). Our approach will apply massively parallel sequencing in a two-stage design, first applying whole-exome sequencing to a familial HNC case resource and then targeted candidate gene resequencing to a well curated HNC case- control series. Thus we propose the following Specific Aims: 1) to prepare and curate a powerful resource of 200 familial HNC cases and then to subject those cases to a whole exome sequencing based search for candidate HNC susceptibility genes, 2) to apply case-control mutation screening to the candidate HNC predisposition genes identified in Aim 1 in a series of 2,000 HNC cases and 2,000 cancer-free controls, and 3) to assess whether the risks conferred by the HNC susceptibility genes differ by sub-site and epidemiologic risk groups.
In specific aim 1, we will compare our results with control exomes from ~500 individuals from the 1,000 Genomes project. We expect to define approximately 500 candidate genes, which will be sequenced in specific aim 2 with a targeted exon capture panel directed at the protein coding exons, promoters, and known transcriptional regulatory elements. The combination of Aims 1 and 2 will have >80% power to reach genome- wide significance (p<2.5x10-6) for genes that have odds ratio and pathogenic sequence variant frequency spectra similar to those of ATM and CHEK2 in breast cancer.
For specific aim 3, we will conduct stratified analyses by HNC sub-site, tobacco smoking, alcohol drinking and HPV 16/18 infection status. We will also assess potential interactions between genes and epidemiologic risk factors on HNC risk. Overall, the study has a strong probability of finding a substantial fraction of the clinically relevant "missing heritability" that contributes to HNC predisposition and of providing key data required to translate these findings directly to clinical practice.
Each year in the US, approximately 53,600 patients are diagnosed with head and neck cancer. Very few head and neck cancer susceptibility genes have been identified thus far. In this project, we will apply newly developed mutation screening technologies to head and neck cancer patients who have a family history of head and neck cancer, to identify new head and neck cancer genes.