Abstract P5-09-02: Breast cancer risk prediction using a polygenic risk score in women of African ancestry: Findings from GWAS in breast cancer in the African diaspora

Abstract

Background: Multiple common susceptibility loci for breast cancer (BC) have been identified/confirmed in Caucasian women. Combination of these SNPs into a polygenic risk score (PRS) could improve risk stratification and provide guidance for preventive and screening strategies. However, due to differences in allele frequencies of genetic variants, tumor characteristics between women of African and European ancestries, we sought to evaluate the association of PRS with BC in a large consortium of African women. Methods: The GWAS in BC in the African Diaspora (ROOT consortium) included 3686 participants of African ancestry from Nigeria, USA, and Barbados (1657 cases, 2029 controls). PRS was constructed from the published odds ratios (ORs) from 90 susceptibility loci for BC. Logistic regression was used to examine its association with overall BC risk as well as associations by hormone receptor status, family history and other clinical features. Results: One unit change in the PRS was associated with an OR of 1.13 (95% CI: 1.01-1.28, P=0.042) for overall BC risk, 1.15 (95%CI: 0.95-1.41, P=0.160) for ER+ BC risk, and 1.17 (95%CI: 0.95-1.44, P=0.133) for ER- BC risk. The ORs for developing BC by percentiles of the PRS, relative to women in the middle quintile, showed weak linear trend. The discriminative accuracy of the PRS, as measured by the C-statistic, was 0.524 (95% CI: 0.505-0.542) for overall BC, 0.511(95% CI: 0.479-0.543) for ER+ BC, and 0.513 (95% CI: 0.481-0.545) for ER- BC. There was a statistically significant interaction between PRS and age, the association between PRS and overall BC risk were stronger in two age groups (aged <40 years and ≥60 years). The PRS was also more strongly associated with PR+ (OR=1.26, 95%CI: 1.01-1.58) compared to PR- (OR=1.08, 95%CI: 0.95-1.44) BC. Association between PRS and BC were similarly non-significant across different strata of family history of BC, BMI, alcohol consumption, oral contraceptive use, menopausal, ER and HER2 status. Table 1. Performance of the 90-SNP polygenic risk score (PRS) in the GWAS in Breast Cancer in the African DiasporaPercentile of PRS (%) OR (95%CI) Overall BC (n=1657) vs Ctrl (n=2029)ER+ BC (n=403) vs Ctrl (n=2029)ER- BC (n=374) vs Ctrl (n=2029)<51.01 (0.73-1.41)1.32 (0.28-2.13)0.89 (0.51-1.55)5-100.83 (0.59-1.15)0.71 (0.40-1.26)0.63 (0.34-1.17)10-200.74 (0.57-0.95)0.69 (0.44-1.07)0.72 (0.46-1.12)20-400.97 (0.79-1.19)0.80 (0.56-1.14)0.98 (0.69-1.39)40-601.00 (ref)1.00(ref)1.00(ref)60-800.89 (0.73-1.10)0.94 (0.66-1.33)0.90 (0.63-1.28)80-901.14 (0.89-1.47)0.87 (0.56-1.34)0.90 (0.63-1.28)90-951.14 (0.82-1.58)1.54 (0.92-2.58)0.88 (0.48-1.61)>951.10 (0.79-1.52)1.15 (0.67-1.99)1.34 (0.79-2.25)Note: Odds ratios are for different percentiles of the PRS relative to the middle quintile (40% to 60%). Odds ratios were adjusted for study site and the first ten eigenvectors from principal components analysis. Conclusion: BC PRS obtained from prior GWASs conducted in Caucasian women didn't provide a comparable degree of risk stratification for African Americans. Additional studies are needed to identify SNPs specific to women of African ancestry that could provide improved risk prediction. Further studies can also combine the PRS with lifestyle/environmental factors. Citation Format: Wang S, Qian F, Zheng Y, Ogundiran T, Ojengbede O, Zheng W, Blot W, Nathanson KL, Hennis A, Nemesure B, Ambs S, Olopade OI, Huo D. Breast cancer risk prediction using a polygenic risk score in women of African ancestry: Findings from GWAS in breast cancer in the African diaspora [abstract]. In: Proceedings of the 2016 San Antonio Breast Cancer Symposium; 2016 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2017;77(4 Suppl):Abstract nr P5-09-02.