Genome-wide association study identifies a novel breast cancer susceptibility locus


Session type:

Gillian Ross1, Olivia Fletcher2, Nicola Johnson2, Nick Orr2, Fay Hoskings2, Lorna Gibson3, K Walker2, Diana Zelenika5, Stephen Chanock6, S Heath1, David Hunter5, Ben Coupland2, Peter Broderick2, Minouk Shoemaker2, Michael Jones1, Sarah Chilcott-Burns1, Mark Lathrop7, Richard Houlston2, Alan Ashworth2
1Royal Marsden Hospital, London, United Kingdom,2Institute of Cancer Research, London, United Kingdom,3London School of Hygiene and Tropical Medicine, London, United Kingdom,4Welcome Trust for Human Genetics, Cambridge, United Kingdom,5Molecularand Genetic Epidemiology, Harvard, United States,6National Institure of Health, Bethesda, United States,7Centre National de Genotypage, Paris, France


Many breast cancers arise in a genetically susceptible minority of women, most of whom are not carriers of mutant BRCA1 or BRCA2. The excess familial risk unaccounted for by BRCA1 qand BRCA2 can be be explained by a polygenic model in which a large number of "low penetrance" variants that confer small risks act in combination to cause wide variation in risk in the population


We conducted a genome-wide association study to identify additonal common variants influencing the risk of developing breast cancer. To enhance the power of our study a high proportion of our cases had 2 primary breast cancers or a family history.We examined 296,114 tagging SNP's in 1769 cases and 2365 controls with validation in 3 independent series totalling 11,955 cases and 12,487 controls.


Combining stage1 , CGEMS and stage 2 data markers in 6 of loci identified in previous GWA studies (2q35, 3p24.1(SLCaA7), 5p12,8q24.21,10q26.13(FGFR2), and 16q12.1(TOX3) were globally significant (P<5x10-7).We identified a novel risk locus for breast cancer on chromosome 9q(odds ratio 0.89, p=1.75x 10-10) and variants mapping to 6q25.1(ESR1) a region that has previously reported as a risk locus in a Chinese population.We also identified a variant mapping to 10q26 that appears to act independently of the published locus within the second intron of FGFR2.


The power of our study to detect the major common loci conferring risks of 1.2 or greater(such as the 10q26 variant) was high(>90%).  There are unlikely to be many additional SNP's with similar effects for alleles with frequencies>0.2 in populations of European ancestry. In contrast we had <50% power to detect alleles conferring relative risks of <1.15 with allele frequencies <0.2. Fine mapping will be needed to identify causal variants and to determine their functional effects.