The ability to establish genetic risk models is critical for early identification and optimal treatment of breast cancer. For such a model to gain clinical utility, more variants must be identified beyond those discovered in previous genome wide association studies (GWAS). This is especially true for women at high risk because of family history, but without BRCA1/2 mutations.

This study incorporates three datasets in a GWAS analysis of women with Ashkenazi Jewish (AJ) homogeneous ancestry. Two independent discovery cohorts were comprised of 239 and 238 AJ women with invasive breast cancer or preinvasive ductal carcinoma in situ and strong family histories of breast cancer, but lacking the three BRCA1/2 founder mutations, along with 294 and 230 AJ controls, respectively. An independent, third cohort of 203 AJ cases with familial breast cancer history and 263 healthy controls of AJ women was used for validation.

A total of 19 SNPs were identified as associated with familial breast cancer risk in AJ women. Among these SNPs, 13 were identified from a panel of 109 discovery SNPs, including an FGFR2 haplotype. Additionally, 6 previously identified breast cancer GWAS SNPs were confirmed in this population. Seven of the 19 markers were significant in a multivariate predictive model of familial breast cancer in AJ women, 3 novel SNPs [rs17663555(5q13.2), rs566164(6q21), and rs11075884(16q22.2)], the FGFR2 haplotype, and 3 previously published SNPs [rs13387042(2q35), rs2046210(ESR1), and rs3112612(TOX3)], yielding moderate predictive power with an area under the curve (AUC) of the ROC (receiver-operator characteristic curve) of 0.74.

Population-specific genetic variants in addition to variants shared with populations of European ancestry may improve breast cancer risk prediction among AJ women from high-risk families without founder BRCA1/2 mutations.