NEW YORK – Researchers at Vanderbilt University and elsewhere have identified genetic loci with ties to breast cancer in women of African ancestry, leading to a polygenic risk score (PRS) with improved performance in such women and new causal clues to the disease.
"Females of African ancestry have been largely underrepresented in studies of germline susceptibility to cancer," senior and corresponding author Wei Zheng, an epidemiology researcher the Vanderbilt University Medical Center's Vanderbilt Epidemiology Center, and his colleagues wrote in Nature Genetics on Monday, arguing that the "present study represents a substantial step forward in reducing those disparities."
Although common variants at some 200 loci have been linked to breast cancer risk in prior genome-wide association studies, the team explained, such research has mainly focused on European ancestry cases and controls. Moreover, only a subset of the breast cancer-associated variants found in non-African individuals showed ties to breast cancer in African ancestry women, suggesting more research is needed to untangle germline genetic contributors to the disease within and across ancestry groups.
"Because of substantial differences in genetic architecture between populations of African and European or Asian ancestry and the limited sample sizes of previous studies on populations of African ancestry, to date, less than 20 percent of breast cancer risk variants identified in the GWAS of descendants of individuals of Asian or European ancestries have been directly replicated in females of African ancestry," the researchers wrote.
With that in mind, the current study centered on tens of thousands of African ancestry individuals with or without breast cancer, who were enrolled through the African Ancestry Breast Cancer Genetic consortium.
The researchers began by bringing together high-density array-based genotyping data or whole-genome sequencing data for 18,034 African ancestry women with breast cancer and more than 22,100 unaffected control participants from the same population.
In the process, they flagged breast cancer variants at 13 known risk loci and unearthed new, genome-wide significant associations at 12 loci, including five loci that were overrepresented in the 9,304 estrogen receptor-positive breast cancer cases and three loci with closer ties to both estrogen receptor-negative breast cancer and triple-negative breast cancer (TNBC), an aggressive breast cancer subtype that is overrepresented and develops at a younger-than-usual age in West African ancestry groups.
In particular, the team highlighted a missense, low-frequency ARHGEF38 gene variant that was associated with breast cancer overall, along with a relatively common TNBC-associated variant called rs76664032 on chromosome 2.
The researchers went on to investigate these and other apparent risk loci in more detail with the help of expression quantitative trait loci found using RNA sequencing-based gene expression data on normal breast tissue samples from 115 women of African ancestry. Meanwhile, their fine-mapping analyses led to 184 credible causal variants.
Across the subset of 2,860 women who had TNBC, the researchers demonstrated that more than 15 percent showed particularly high genetic risk of the condition, carrying at least six TNBC-related alleles at sites linked to the subtype in the past. They estimated that breast cancer risk rose more than fourfold in African ancestry women carrying six risk alleles relative to that found in those who had two risk alleles or fewer.
Likewise, an 89-variant PRS trained using data for 15,680 African ancestry women with breast cancer and 17,362 without showed improved performance for distinguishing 2,354 breast cancer cases from 4,742 controls compared to a PRS containing 236 risk variants from a breast cancer PRS based on European ancestry studies.
"The PRS derived in our study specifically for females of African ancestry had breast cancer prediction approaching the level observed previously for other populations," the authors reported, adding that the work also "substantially adds to population diversity in current genetic studies."