The contribution of common genetic variants to breast cancer risk in South African black populations

Abstract

Breast cancer is the second most common cancer in South African black women. The contribution of common genetic variants to breast cancer risk is well studied in non-African populations, but little is known about their role in resident African populations, and there are no published genome-wide association studies (GWAS) on breast cancer in Africa. This PhD thesis aimed to determine the contribution of common genetic variants to breast cancer in a South African black population. A GWAS was carried out in 2,573 black female breast cancer patients from the Johannesburg Cancer Study and 744 population-matched, female controls from the AWI-Gen study. All participants were from Soweto, Johannesburg, South Africa. Samples were genotyped on the H3Africa SNP array. Replication testing was done of existing loci from European and African American (AA) populations in the resident African data, and loci from the resident African data in European and AA populations. A meta-analysis was carried out with an AA population. Finally, existing polygenic risk scores (PRSs) were tested in the resident African dataset. Three variants at two loci were strongly associated with breast cancer in this study. Two variants (rs77422433, p-value=2.89x10-08, odds ratio (OR):0.46, 95% confidence interval (95%CI): 0.40-0.52 and rs112410019, p-value=3.01x10-08, OR: 0.47, 95%CI: 0.41-0.53) were located within the DNA repair gene XRCC5. These variants were not previously associated with breast cancer, suggesting that it may be an African specific risk locus. The second locus is on chromosome 16 in CES5A (rs3859109, p-value=4.54x10-08, OR=0.70, 95%CI: 0.68-0.73), and had not previously been associated with breast cancer. None of these SNPs were replicated in European and AA populations. The meta-analysis with AA data revealed strong association of an intergenic SNP with breast cancer (rs139299680, pmeta=7.25x10-08) on chromosome 3. A polygenic risk score (PRS) developed in European populations demonstrated poor transferability to this African dataset. This GWAS is the first to be conducted in a resident black African population. This study suggests that there may be African-specific genetic risk factors for African breast cancer, and that large genome-wide studies in African populations are essential to develop a comprehensive understanding of the genetics of breast cancer in Africa.