10 May 2010
Breast cancer is the most common cancer in women in the UK with around 45,700 women diagnosed in 2007 (according to the leading charity Cancer Research UK). A new study published online in the journal Nature Genetics has identified five novel breast cancer susceptibility loci taking the number of identified genetic variants associated with susceptibility up to 18 [Turnbull et al. (2010) Nature Genetics 9 May; doi:10.1038/ng.586].
Turnbull et al. conducted a genome-wide association study (GWAS) in which almost 600,000 SNPs were genotyped in more than 3,500 breast cancer cases with a positive family history and nearly 5,000 controls. In addition to finding associations with all 13 of the previously identified loci, a further 28 SNPs in 13 regions were also identified. Once linkage disequilibrium was accounted for, 15 SNPs were assessed in the replication sample of over 12,500 cases and over 12,000 controls. Five of these SNPs from five loci on chromosomes 9, 10, and 11 (three loci on chromosome 10) showed evidence of replication with statistically significant results (with P-values ranging from 4.6 x 10-7 to 3.2 x 10-15).
Four of the five SNPs showed stronger association for oestrogen receptor-positive breast cancer cases with little association with oestrogen receptor-negative cases, although this is a pattern observed with the majority of breast cancer loci identified thus far. The authors identified 13 plausible candidate genes across these loci with some genes also showing associations with other cancers (such as malignant melanoma or pancreatic cancer). These five loci when combined with the 13 other common disease susceptibility loci help explain approximately 8% of the familial risk of breast cancer (with a further 20% explained by the rarer mutations in BRCA1 and BRCA2).
Comment: This well conducted GWAS provides evidence for an additional five common breast cancer susceptibility loci in what is the largest breast cancer GWAS conducted to date. However, these novel loci are only estimated to account for an additional 1.2% of familial risk. When taken with the other known common and rare associated loci, around 30% of familial risk is explained with the authors emphasising that the remaining “familial risk is therefore likely to be due to a combination or a large number of common variants with smaller effects together with rarer variants not testable with current arrays” along with epigenetic and other shared environmental factors.
Although unlikely to improve the usefulness of predicting genetic risk at an individual level, the newly identified susceptibility variants could prove beneficial for stratifying the general female population for more effective targeting of screening programmes to those at greatest risk of breast cancer (see previous news). The PHG Foundation is a partner in a large European consortium, the Collaborative Oncological Gene-environment Study (COGS), a project that is developing the understanding of both genetic and environmental risk factors for breast, ovarian and prostate cancers, for use in individual risk prediction.