Performance of Polygenic Risk Scores for Breast Cancer in African Populations: A Systematic Review and Meta-Analysis
This systematic review and meta-analysis reveals that while polygenic risk scores for breast cancer show modest predictive performance in African populations, European-derived models currently outperform those developed specifically for African ancestry, underscoring the critical need for greater genomic representation and recalibration to improve clinical utility.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Breast cancer remains a leading health challenge for women worldwide, and while age is a known factor, a person's genetic makeup also plays a significant role in determining their risk. For decades, scientists have looked for specific changes in DNA, known as single nucleotide polymorphisms, that act as small markers for disease. Individually, these markers offer only a tiny hint of risk, but when combined, they can form a cumulative picture of a person's susceptibility. This combination is called a polygenic risk score. It functions like a personalized genetic report card, intended to help doctors identify women who might need earlier or more frequent screening. However, the vast majority of these genetic reports have been written using data from people of European ancestry. Because human genetic diversity is greatest in Africa, and because the patterns of DNA vary significantly across the globe, there is a growing concern that these existing scores may not work correctly for women of African descent.
A team of researchers set out to test this concern by gathering every available study that had tried to use these genetic risk scores on women of African ancestry. They conducted a systematic review, a rigorous method of collecting and analyzing all relevant scientific papers on a topic, to see how well these tools actually performed. The team searched through four major scientific databases, looking for studies that measured the ability of these scores to distinguish between women who had breast cancer and those who did not. They focused specifically on a statistical measure called the area under the curve, which acts as a gauge for accuracy. A score of 0.50 would mean the tool is no better than a coin flip, while a score of 1.00 would mean it is perfect. The researchers examined fourteen studies that met their strict criteria, involving thousands of women and hundreds of thousands of genetic markers.
The results painted a clear, albeit modest, picture. When the researchers combined the data from all fourteen studies, the overall accuracy of these polygenic risk scores for women of African ancestry was 0.57. This indicates that the tools currently available have only a limited ability to predict who will develop the disease in this population. The researchers found that the scores performed slightly better when they were recalibrated or adjusted for the specific genetic background of the people being tested, but even the best-performing models fell short of the high accuracy seen in European populations. One of the most common scores, which uses 313 genetic markers, achieved an accuracy of 0.59 in African populations, which is better than the average but still far from the level needed for reliable, standalone medical use.
A deeper look into the data revealed why these scores struggle. The researchers found that the accuracy of a score depends heavily on the population used to create it. Scores developed using data from European populations performed poorly when applied to African women, largely because the genetic patterns in Africa are more complex and diverse. Surprisingly, the few scores that were built specifically using data from African populations performed even worse, with an accuracy of 0.53. The authors suggest this is not because African genetics are harder to understand, but because there simply have not been enough large studies involving African women to build a robust model. The data available so far is too small to capture the full range of genetic risk factors unique to the continent.
The study also highlighted that the quality of the research itself matters. When the researchers looked only at the studies with the most rigorous methods and the least chance of error, the accuracy of the scores improved slightly, but the fundamental limitation remained. They found that the specific mix of genetic markers used in each score varied widely, yet there was a slow trend toward using a similar set of markers across different studies. One specific genetic marker, located in a gene called EBF1, appeared in every single study included in the review, suggesting it plays a universal role in breast cancer risk. However, the differences in how these markers behave in different populations mean that a score created for one group cannot simply be copied and pasted for another.
Ultimately, this review concludes that while polygenic risk scores are a promising tool, they are not yet ready to be used as a standalone method for predicting breast cancer risk in African women. The current tools, which were built on European data, do not translate well to the genetic diversity found in Africa. The researchers emphasize that the path forward requires generating large, high-quality genetic datasets specifically from African populations. Until such data exists, the most effective approach may be to take existing scores and carefully adjust them for local populations, while also combining genetic information with other known risk factors like family history and lifestyle. The goal is to ensure that the benefits of precision medicine are shared equitably, rather than leaving a significant portion of the world's population behind.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.