In the news

According to Cancer Research UK, one in nine women in the UK will develop breast cancer at some point in their life. Some major genes involved in rare inherited (familial) forms of breast cancer are well known, and a genetic test for these BRCA1/2 mutations is already available from Myriad Genetics (see previous news), they only account for a tiny proportion of breast cancer cases. ‘Normal’ or ‘sporadic’ breast cancer is believed to arise from a combination of many different interacting environmental and genetic factors, and some relatively common genetic factors have been identified as being linked to a slightly increased lifetime risk of developing breast cancer.

 

The Icelandic company deCODE has now launched a new genetic test to screen for risk of the most common forms of breast cancer (see deCODE news). The deCODE BreastCancer™ test is based on seven common single nucleotide polymorphisms (SNPs), which are each associated with a small relative risk. The results from all seven SNPs are combined to estimate a woman’s risk of breast cancer relative to the population. The company claim that the test can identify those women whose relative is 1.66 or higher, representing approximately 5% of the total female population. They suggest that such women are at sufficiently high absolute risk to meet the American Cancer Society’s recommendation for regular breast screening by MRI.

 

Comment: The test has drawn heavy criticism from a number of commentators, who assert that the test clinically useless and could lead to women be either unduly alarmed or falsely reassured, although there is currently little or no evidence for either of these outcomes. There is a valid concern that, in the absence of professional medical advice and proper counselling, members of the public may misinterpret the test, which is currently available both direct-to-consumer as part of a larger genome-wide scan for risk of numerous diseases, and through physician referral specifically for breast cancer risk . But what are the other potential problems?

 

There are a number of points to consider when assessing a test of this sort. Firstly, the legitimacy of the test – does it measure what it claims to measure, and is there strong scientific evidence linking the genetic variant to the disease? In this case, the former question is addressed by the fact that the assay is carried out in a CLIA-waived laboratory, and is therefore likely to be accurate, and the latter by a number of large studies indicating a real association between each of the seven SNPs and breast cancer. The relative risk estimate generated by the test are valid and so the test itself is scientifically legitimate. However, the exact nature of the claims and the usefulness of the test may not be.

 

The second question is therefore one of interpretation – what do the results of the test actually mean for an individual patient? The test places an individual within a sub-population with a average specific risk, but cannot locate the exact whereabouts of that individual within the distribution of risks of the whole group. The key issue in interpretation is the conversion of relative to absolute risk. However, this task is not simple, as it is age dependent and also depends on whether competing mortality is taken into account (which many published absolute risks do not). For example, the average absolute risk that a woman who has not had breast cancer will develop the disease before the age of 85 ranges from around 8% (1 in 12.5) at ages 20 or 30, through 6.5% (1 in 15) at age 50 to 3% (1 in 33) at age 70.

 

Finally, there is a question of utility – do the benefits of the test outweigh the potential harms, and can the test usefully provide guidance for action? These questions are not simple to answer, and will depend enormously upon the context in which the test is used, and ultimately upon the purpose of testing. If the purpose, as is suggested by some commentators, is to accurately inform a patient whether they will, or will not, develop breast cancer, then it is unlikely to be clinically useful. However, if the purpose is to identify women at a higher than average risk in order to target them for screening, as suggested in a recent paper in the New England Journal of Medicine (see previous news), then it could potentially be very useful.

 

However, even this final point is fraught with difficulties, as the guidance on when to offer screening (or indeed any form of intervention) is highly controversial and varies between countries. Although the American Cancer Society recommends annual mammography from age 40, the UK NHS Breast Screening Programme is only offered to women aged 50-70, who are at a substantially higher risk of developing breast cancer within the following 10 years. Determining whether this type of test will ultimately yield tangible benefits through a reduction in the breast cancer morbidity and mortality will take much more research, a long time and a lot of data.

 

Therefore, although it is certainly too early to offer such a test to women routinely, and proper provision for professional medical support should always be available, our suggestion is that providing the test in the marketplace is not in itself illegitimate. How the test is to be used, how it is to be interpreted and what interventions are offered to a woman with higher risk, however, are questions that deserve much greater scrutiny.