Skip to main content

27-07-2010 | Oncology | Article

Polygenic risk score predictive of breast cancer subtype


Free abstract

MedWire News: A study of common genetic variants for breast cancer has revealed that a polygenic risk score is substantially more predictive of estrogen receptor (ER)-positive than of ER-negative breast cancer.

Gillian Reeves (University of Oxford, UK) and colleagues conducted the study to assess the risk for breast cancer, both overall and by tumor subtype, according to the expression of 14 individual single-nucleotide polymorphisms (SNPs) and a polygenic risk score.

The study included 10,306 women with breast cancer, diagnosed at a mean age of 58 years, and 10,393 women without breast cancer who had participated in a prior genotyping trial. Reeves and team estimated the per-allele odds ratio (OR) for breast cancer of individual SNPs and the cumulative incidence of breast cancer up to the age of 70 years, according to a polygenic risk score based on the four, seven, or 10 SNPs most strongly associated with developing breast cancer.

The researchers found that the risk for breast cancer was greatest for the SNPs FGFR2-rs2981582 (OR=1.23) and TNRC9-rs3803662 (OR=1.20) and that these two SNPs were more strongly associated with ER-positive than ER-negative disease.

The next strongest association was for the SNP 2q-rs13387042, for which the per-allele OR was significantly greater for bilateral than unilateral disease (OR=1.39 vs 1.15), and for lobular than ductal tumors (OR=1.35 vs 1.10).

The estimated cumulative incidence of breast cancer to age 70 years among women in the top compared with bottom fifth of a polygenic risk score based on seven SNPs was 8.8% versus 4.4%, respectively. This difference was much greater for ER-positive disease, at 7.4% versus 3.4%, than for ER-negative disease, at 1.4% versus 1.0%. The findings did not differ significantly when four or 10 SNPs were included in the polygenic risk model.

Other established risk factors, notably a family history of breast cancer, have similar or even greater effects on breast cancer incidence than those observed in this study, Reeves and co-authors remark in the Journal of the American Medical Association.

"Hence, as others have suggested, subdividing women on the basis of their polygenic risk is, at this stage, not a useful tool for population-based breast cancer screening programs," the researchers note. They add, however, that such risk scoring "may be useful for understanding disease mechanisms.

MedWire ( is an independent clinical news service provided by Current Medicine Group, a trading division of Springer Healthcare Limited. © Springer Healthcare Ltd; 2010

By Laura Dean

Related topics