Evidence
View commentaries on this research

This is a plain English summary of an original research article

Genetic tests which predict the risk of breast cancer in White Europeans need to be adapted for use in women from other ethnic groups. New research found that these risk scores are inaccurate and exaggerate risk in Black, Asian, mixed-race and Ashkenazi Jewish women.

The risk of developing breast cancer is strongly influenced by genes. Researchers have identified many versions (variants) of genes that each slightly increase or decrease the risk of breast cancer. They have developed assessments called polygenic risk scores (PRS) which combine the risks of these genetic variants.

These scores tend to be derived from studies of White European women; they effectively predict risk in this group. But new research found that they exaggerate the risk for women in other ethnic groups and could lead to unnecessary worry and interventions. Researchers say there is an urgent need to develop risk scores appropriate for specific ethnic groups.

This study looked at the accuracy of two established risk scores – called SNP18 and SNP143. These tests were accurate in predicting breast cancer in White European women. But they were not accurate in Black, Asian, mixed-race and Ashkenazi Jewish women. Overall, women in these groups who did not have breast cancer still had risk scores well above average. One of the tests (SNP143) overestimated the risk by 40% on average, in women of non-White European origin.

Current NICE guidelines [CG164] recommend that risk models are used alongside standard risk factors such as age or family history. This research suggests that risk scores should only be used in women of White European origin.

Further information on predictive genetic tests for cancer risk genes can be found on the NHS website.

What’s the issue?

Many factors combine to determine someone’s risk of developing breast cancer. Genetic variants and standard risk factors such as age and physical activity can all play a part. Someone with a family history of breast cancer may be referred for further tests to better understand their risk of developing the disease. They are sometimes offered extra screening, drugs to prevent breast cancer or surgery to remove breasts (mastectomy).

Polygenic risk scores combine the risk from several genetic variants associated with breast cancer. To date, most of these genetic variants have been identified in large population studies comparing the genetic make-up of women of White European origin with and without breast cancer.

Genetic variants differ across populations. Risk scores developed in White European women may therefore be unsuitable for women in other ethnic groups.

In the current study, researchers assessed the accuracy of two established risk scores (SNP18, SNP143) for breast cancer prediction in women who were not of White European origin.

What’s new?

The Predicting Risk of Cancer At Screening (PROCAS) study ran between 2009 and 2015 and included women aged 46-73 years. Participants were asked about their ethnicity (Asian/Asian British, Black/Black British, Mixed, White British, Jewish or other) in a questionnaire.

This study looked at the performance of the two tests (SNP143 and SNP18) in women who had consented to genetic (DNA) analysis of a saliva sample. It compared the results in 645 women who developed breast cancer during the study, with 1,868 women who did not.

These analyses found that:

  • in White European women, both tests (SNP18 and SNP143) discriminated well between those who developed breast cancer and those who did not; the scores effectively predicted the risk of breast cancer
  • in women of other ethnic groups (112 Black, 119 Asian, 44 mixed and 120 Ashkenazi Jewish women), both tests overestimated the risk of breast cancer; this was most marked for Black and Jewish women
  • the SNP143 test overestimated the risk of breast cancer by 40% in the combined group of non-White and non-European women; SNP18 overestimated it by 14%.

Why is this important?

These findings demonstrate that polygenic risk scores developed in one ethnic group can give misleading results in other populations.  An overestimate of risk could provoke unnecessary anxiety and lead to excessive screening or preventative drug and surgical treatments.

The scale of the overestimated risk in this study suggests that wider use of polygenic risk scores is likely to increase health inequity. Women from non-White European backgrounds cannot rely on the tests available at present.

What’s next?

An immediate first step is for all tests to be clearly labelled with the ethnicities they have been developed and tested in. National breast cancer detection programmes in the US and Europe need to take ethnicity into account when using the tests.

Clear labelling is especially important as the tests become commercially available. Currently, companies do not always make clear that the tests can be inaccurate for women from different ethnic groups. Some tests state that they are only valid for White European and Ashkenazi Jewish women. However, this study found the tests overestimate risk in Ashkenazi Jewish women.

Tests are being developed for Asian and Hispanic populations and, ideally, there will be different tests for each ethnic group. However, there is insufficient research among Black populations. British Black (predominantly Afro-Caribbean) populations are likely to differ from African populations. The challenge of recruiting sufficient numbers of participants will need to be overcome, the researchers say.

You may be interested to read

The paper this NIHR Alert is based on: Evans DG, and others. The importance of ethnicity: Are breast cancer polygenic risk scores ready for women who are not of White European Origin? International Journal of Cancer 2022;150:1

A study looking at breast cancer risk during screening: information about the PROCAS study from Cancer Research UK.

National Institute for Health and Care Excellence (NICE). Familial breast cancer: classification, care and managing breast cancer and related risks in people with a family history of breast cancer. NICE guideline [CG164]. 2013

Information on risk factors for breast cancer from Cancer Research UK.

 

Funding: This study was supported by the NIHR Manchester Biomedical Research Centre.

Conflicts of Interest: One of the study authors provides consultancy to Myriad Genetics. Two have received royalty payments through Cancer Research UK for the use of a breast cancer risk assessment algorithm.

Disclaimer: NIHR Alerts are not a substitute for professional medical advice. They provide information about research which is funded or supported by the NIHR. Please note that views expressed in NIHR Alerts are those of the author(s) and reviewer(s) and not necessarily those of the NHS, the NIHR or the Department of Health and Social Care.


  • Share via:
  • Print article

Commentaries

Study author

These findings show you can’t simply apply polygenic risk scores (PRS) derived from White Europeans in people of other ethnicities. Black, Asian, mixed-race and Jewish women in the UK might buy commercial tests that could exaggerate their risk of the disease. This may frighten them and lead to unnecessary screening and other interventions they don’t need.There is an urgent need for more research to develop ways of adjusting these risk scores for different ethnicities. We need more accurate risk predictions to avoid further increasing existing health disparities in minority populations in the UK.Gareth Evans, Professor of Medical Genetics and Cancer Epidemiology, the University of Manchester 

Prevent Breast Cancer 

Prevent Breast Cancer are delighted to support studies such as this one, that demonstrates that there is vitally important work still to be done to ensure equity of access to information about the risk of developing breast cancer across all ethnic groups.Nikki Barraclough, Executive Director of Prevent Breast Cancer 

Researcher  

Polygenic risk scores (PRS) indicate a person’s risk of diseases using knowledge of what genetic variants are linked with disease. Regrettably, most of this data is from White participants as the numbers of non-White participants in studies are often low, leading to their exclusion from analyses.Genetic differences due to ethnicity, and sex, are becoming better understood in the research community. However, as the authors state, this is less well-known among clinicians and the public and current PRS can overestimate risk for those from non-White backgrounds..In the UK, people are more likely to encounter PRS via private companies. These companies may not provide information on the compatibility of their score with consumer ethnicities or disclose which populations their score is based on. Lack of contact with a healthcare professional after receiving results could lead to undue harm and worry in those from minority ethnic groups whose risk may be overestimated. These communities, alongside healthcare professionals, should be made aware of the potential bias in current PRS to mitigate this potential harm.Greater national and international collaboration, and a focus on non-White communities in genetic studies, is needed to remove PRS bias, especially if PRS is to be adopted within the NHS.Sarah Harbach, PhD Student, School of Biological Sciences, University of Manchester 

Member of the public

This research should improve the quality of professional advice and lead to more rational use of NHS resources. Commissioning for healthcare needs to be based on sound research. The findings should be incorporated into public health screening policy and into the evidence base for professional advice.I would like to see more collaboration with similar research from other locations, such as from Singapore, Korea, Japan, India, Pakistan, Taiwan, Africa and Hong Kong, That would further add to our understanding.Emily Lam, Public Contributor, Macclesfield 

Member of the public

I assume most health care professionals will be more cautious about using PRS in non-White, non-European women. I hope that healthcare guidelines will also reflect this caution. Women and clinicians working in the field need to be aware of these results.However, we need further research on much larger groups of non-White, non-European women.Cathie Joyce, Public Contributor, Shrewsbury 
Back to top