New Breast Cancer Risk Models for Women of African Ancestry
Despite advancements in genetic testing for breast cancer, death rates remain elevated among women of African ancestry. This disparity is often linked to the inaccuracy of existing genetic models, higher prevalence of aggressive tumor subtypes, and later diagnoses.
Researchers at the University of Chicago Medicine have developed new polygenic risk score (PRS) models to enhance breast cancer risk prediction for women of African ancestry. These models, utilizing genetic data from over 36,000 women, represent a comprehensive tool for this population. The findings were published in Nature Genetics.
Limitations of Existing Models
Most current genetic tools for breast cancer risk prediction were developed using data primarily from women of European ancestry. While effective for that group, these models often fail to provide accurate predictions for women of African ancestry, particularly for aggressive subtypes like triple-negative breast cancer (TNBC).
This discrepancy is attributed to genetic diversity. Populations of African ancestry exhibit greater genomic variation, and differences in the frequency and distribution of genetic variants can influence disease patterns.
Risk models based on European genetic data may miss critical genetic signals present in African DNA.
Polygenic risk scores assess a patient's DNA for single nucleotide polymorphisms (SNPs). A higher number of specific SNPs can increase cancer risk. According to Dezheng Huo, PhD, Professor of Public Health Sciences and senior author, previous PRS models were less accurate for African American women due to smaller sample sizes and greater genetic diversity. A large consortium was formed to improve prediction accuracy.
Model Development and Performance
The research team, led by Huo, developed new PRS models specifically for women of African ancestry using data from the African Ancestry Breast Cancer Genetics Consortium, which includes women from the U.S., the Caribbean, and Sub-Saharan Africa. The models were created for overall breast cancer, estrogen receptor positive (ER+), estrogen receptor negative (ER-), and triple-negative breast cancer (TNBC).
Model performance was measured by its area under the curve (AUC), with scores closer to 1 indicating better prediction.
The new models achieved AUC scores between 0.61 and 0.64, a significant improvement over previous models that scored in the 0.56-0.58 range.
Simplified models were also developed to improve usability and reduce costs; for example, a TNBC model using 162 genetic markers achieved an AUC of 0.626.
Improved risk prediction can enable earlier screening and tailored care. The study suggests that women in the top 1% of risk scores could benefit from screening as early as age 32, which is earlier than current recommendations.
Impact of Combining PRS with Family History
When combined with family history, the new PRS models show even stronger predictive capabilities.
Women in the top 1% of PRS scores who also had a first-degree relative with breast cancer exhibited a lifetime risk exceeding 50%.
This level of risk could justify earlier and more frequent screening, as well as preventive interventions.
Validation and Future Directions
The models were validated in independent datasets, including the All of Us study, confirming consistent performance across different populations. The study primarily focused on African American and women of West African ancestry, and researchers emphasized the need for further research to account for genetic differences among diverse African populations globally.