Molecular Epidemiology of Breast Cancer

Abstract: Hereditary breast cancer constitutes a considerable fraction of the total number of breast cancer cases occurring each year. Up until recently very few breast cancer predisposing genes were known, but many new common polymorphisms contributing to increased cancer susceptibility are continuously being identified. This thesis has focused on familial breast cancer and identifying as well as investigating common low-penetrant polymorphisms contributing to breast cancer risk. We hypothesized that since methylation of the promoter region is a common phenomenon of tumour suppressor genes, turning them off. Inherited methylation potential, in the form of common CpG-SNPs, might affect cancer risk. We conducted a large study in five different independent population cohorts (comprising totally >3000 cases) to test this hypothesis in genes previously implicated either in breast cancer or methylation. In this study we were able to identify one SNP possibly associated with breast cancer in the ESR1 gene. We also examined previously identified common variants affecting breast cancer risk by replicating them in the same five cohorts and examining how the risk increased with increasing number of risk-alleles. A highly significant increasing trend was seen (p=9.3x10-26). Based on comparisons with other replication studies using different study designs we concluded that the addition of SNPs from the two, highly replicated, loci FGFR2 and TOX3 could add information to screening of high risk families. Possible interactions between common genetic variants and established environmental or phenotypic risk factors for breast cancer were examined in two different studies comprising 2063 and 728 breast cancer cases respectively. The significant findings from these two studies were few and may be contributed to coincidence. In summary we found that methylation potential might be a factor worth considering when searching for SNPs implicated in cancer risk and that common low-penetrant variants contribute to breast cancer risk with the risk increasing substantially with increasing number of risk alleles. To further evaluate possible interactions between common variants and environmental risk factors very large cohorts will be needed.