Genetic information doesn't improve breast cancer risk prediction, large study finds

Breast cancer risk prediction does not improve significantly when genetic information is included in the risk prediction model, a new study published in the New England Journal of Medicine has found.

Researchers identified 10 gene variants - SNPs (single nucleotide polymorphisms) - that have been linked to the increased risk of breast cancer in genome-wide association studies. They asked whether the gene variants could improve the prediction of breast cancer risk over the conventional Gail model that is based on a woman's family, reproductive and medical history. Including genetic risk did not significantly improve breast cancer risk prediction. The authors conclude that for most women, knowledge of their personal genetic information would not change clinical recommendations for breast cancer screening or treatment.

'When we included these newly discovered genetic factors, we found some improvement in the performance of risk models for breast cancer, but it was not enough improvement to matter for the great majority of women,' said Sholom Wacholder, senior investigator at the US National Cancer Institute, National Institutes of Health.

The study included 5,590 breast cancer patients and 5,998 women without cancer, between the ages of 50 and 79, from the NCI Cancer Genetic Markers of Susceptibility genome-wide association study of breast cancer. For each participant, researchers used genotype information of the 10 SNPs associated with breast cancer, as well as information from the Gail model: the number of first-degree relatives with a diagnosis of breast cancer, age at menarche, age at first live birth, and number of prior breast biopsies. They then calculated the probability that a woman who goes on to develop cancer would have a higher risk prediction - based on the combined Gail and genetic risk model - than a woman who did not develop cancer.

The researchers found that risk prediction based on the 10 SNPs alone was as good as the conventional Gail model at predicting breast cancer risk. However, combining the genetic risk and the Gail model did not significantly improve breast cancer risk prediction over the Gail model alone. The changes in risk prediction were too small to influence clinical decision-making.

 

'Our results indicate that the recent identification of common genetic variants does not herald the arrival of personalized prevention of breast cancer in most women. Even with the addition of these common [gene] variants, breast-cancer risk models are not yet able to identify women at reduced or elevated risk in a clinically useful way.' The researchers concluded that, 'given the cost involved, genetic screening is not worthwhile in a clinical context'.

 

In an editorial that accompanies the article, Peter Devilee of Leiden University Medical Center and Matti Rookus of the Netherlands Cancer Institute in Amsterdam, agree that 'for women seeking advice on their personal risk of breast cancer, it is obviously too early to incorporate SNP testing into a counselling procedure, although such tests are already advertised for this purpose on the Internet'.

However, they also stress that the predictive value genome-wide association studies is improving, and they will increasingly be able to detect very uncommon gene variants that might be more strongly associated with breast cancer. They point out that, individually, each of the 10 SNPs used in the study only increase the risk of cancer by a small amount. Furthermore, the study did not include the BRCA-1 and BRCA-2 breast cancer genes that are known to confer a high risk of breast cancer in women with mutations in these genes. Myriad Genetics has patented the BRCA gene tests and hold the exclusive rights to their use.

Devilee and Rookus predict that 'the performance of our tools to stratify women according to their risk of breast cancer will probably increase dramatically over the next decade'. Wacholder and colleagues agree that 'we can expect to identify more genetic determinants of breast cancer, and to learn more about those we have already found. This information, along with our increasing knowledge of non-genetic factors, should allow us to steadily improve our risk prediction models for breast cancer.'