My ePortfolio Register   

Polygenic risk scores can refine breast cancer risk assessment for BRCA negative high-risk women

By ecancer reporter Clare Sansom

Breast cancer is the commonest tumour type in women; the lifetime risk of a woman developing the disease if she lives well into old age in the UK and Europe is about one in eight.

However, breast cancer risk varies widely between individual women, with genetics playing a large part in this.

It is about twice as common in women with at least one first-degree relative (mother, daughter or sister) with the condition as it is in the general population.

The most significant genetic risk factors are mutations in the well-known tumour suppressor genes BRCA1 and BRCA2, but only a minority of high-risk women will carry a mutation in one of these genes.

Many other genes that can influence breast cancer risk have been identified, but the increase in risk associated with each of these and, therefore, their utility as predictors is very low.

Panels of risk-associated genes are likely to be of greater use as predictors than individual genes.

A set of eighteen SNPs, each in a different gene, have been identified as accounting for about two-thirds of the component of familial breast cancer risk not associated with BRCA mutations.

A group of researchers led by D. Gareth Evans from the Manchester Centre for Genomic Medicine, University of Manchester, UK have now conducted a case-control study of a polygenic risk score (PRS) derived from these eighteen variants in predicting breast cancer cases in women with and without BRCA mutations.

All cases of breast cancer in women with a family history of the disease who attended the Genesis Prevention Centre in Manchester between 1987 and 2014 were recorded, and cases and age-matched cancer-free controls who had attended the same clinic asked to provide blood samples for DNA analysis.

Stored samples of DNA extracted from the blood of breast cancer patients who had died from their disease were also included in the study.

The complete cohort included 451 breast cancer cases, 157 of whom carried BRCA mutations, and 1,605 controls (405 with BRCA mutations).

An odds ratio (OR) was calculated for the three possible genotypes – 0, 1 or 2 risk alleles – for each of the 18 SNPs included in the study, and an overall polygenic risk score (PRS) calculated for each woman by multiplying together the eighteen ORs associated with her individual genotypes.

A questionnaire was used to determine non-genetic risk factors, including ages at menarche, first child (if parous) and, if relevant, menopause; height and weight; distant relatives diagnosed with breast and ovarian cancer and history of benign breast disease, and these were incorporated into each woman’s risk score.

The cohort was stratified by BRCA status, and the predictive quality of the combined risk scores evaluated using a univariate likelihood ratio (LR) x2 for risk associated with log PRS.

Overall, PRS was found to be a significant risk factor for predicting breast cancer in the group without BRCA mutations (LR x2 22.7, p<0.001) and, and the findings were similar if the group of patients who had been diagnosed with the disease before attending the clinic (non-prospective cases) were excluded.

A quintile analysis of this group showed that the 20% of women with the highest PRS values were more than twice as likely to develop breast cancer than the 20% with the lowest values.

Adjusting these women’s previously determined risk scores using the PRS moved about half the women between defined risk categories with, for example, 18% of women moving into and 20% out of the 25% lifetime risk group that will be offered MRI scans under current US rules.

Results from the group of women with mutations in BRCA1 or BRCA2 were less significant, partly because the group was smaller, but there was some suggestion that the SNP scores could usefully refine overall risk in these women.

Evans and his co-workers concluded that this score derived from eighteen SNPs with low individual effect sizes should be useful in refining risk calculations for BRCA negative women who attend family history clinics, and that further studies were necessary to determine its utility for those bearing BRCA mutations.


Evans, D.G., Brentnall, A., Byers, H., Harkness, E., Stavrinos, P., FH-risk Study Group, Newman, W.G. and Cuzik, J. (2017). The impact of a panel of 18 SNPs on breast cancer risk in women attending a UK familial screening clinic: a case-control study. J. Med. Genet. 54(2):111-113. doi: 10.1136/jmedgenet-2016-104125.




Please click on the 'New Comment' link to the left to add a new comment, or alternatively click any 'Add Comment' link next to any existing post to respond. The views expressed here are not those of ecancer. For more information please view our Privacy Policy.

Founding partners

European Cancer Organisation European Institute of Oncology

Founding Charities

Foundazione Umberto Veronesi Fondazione IEO Swiss Bridge

Published by

Cancer Intelligence