Polygenic risk scores for breast cancer: ready or not?

Bookmark and Share
Published: 14 Dec 2018
Views: 3352
Prof Gareth Evans - University of Manchester, Manchester, UK

Prof Gareth Evans speaks to ecancer at SABCS 2018 about polygenic risk scores for breast cancer.

He explains that the development of these polygenic risk scores means that they can be used in a model alongside mammographic breast density and other risk factors to accurately predict breast cancer risk and also better stratify for early detection.

He believes that this measure is ready and that it wont be long until it reaches standard practice.

What I talked about yesterday was the use of common genetic variants called SNPs, single nucleotide polymorphisms, in being able to more accurately predict the risk of breast cancer. There are now 313 of these SNPs that contain information that can either reduce or increase your risk of breast cancer. When you put them all together into what’s called a polygenic risk score this has huge power to predict the risk of breast cancer. So we’re now at a stage where we can use this polygenic risk score, alongside other risk factors like the density of the breast tissue, mammographic density, and standard risk factors that have been known for many years, putting those three together in a model means we can accurately predict the risk of breast cancer in the next ten years and potentially for the entire lifetime.

The implication of that is that we can better risk stratify the population for early detection, starting screening earlier in those at greater risk, starting it later in those who are low risk, but also use this to target women who might be interested in taking prevention drugs like tamoxifen, aromatase inhibitors like anastrozole, and also at the extreme end there may be women who actually get pushed into a risk category where risk reducing surgery might be an option.

So the topic of the talk was is this ready?

Yes, it is ready, it’s just a matter of time before this starts to get embedded into standard practice. I would hope in the future that national screening programmes like the National Health Service breast screening programme in the UK will actually use risk stratification to more intelligently screen in the general population such that you’re not over-diagnosing loads of women at low risk and you’re targeting more frequent screening in those at higher risk. We can identify now 18%, one in six of the population, who develop 40% of the cancers and about 45% of the stage 2 cancers. Those can potentially be either prevented or downstaged by more frequent screening.

What do you look for when examining these risks?

What we basically use is an OncoArray that has been designed specifically to identify those SNPs but also a huge number of additional SNPs that may eventually be proven to be linked to breast cancer. We identify the nucleotide that they carry, whether they carry two copies of the risk allele or no copies or one copy, and then you can put that together, normalising the odds ratio around one and you literally just multiply together the score from each SNP and you come out with a polygenic risk score which varies from as low as 0.2, so fivefold reduction in risk, up to about a four- to fivefold increase in risk.

How long do you think it will be until physicians are using this to determine treatment?

There are already private companies that have set this up so it’s part of a number of companies’ genetic risk profiling. It’s only a matter of time; we’re setting this up in Manchester, Manchester has a devolved health service, and we’re in advance plans to actually use this in the general population to target women in their mid-50s who might be eligible to take anastrozole and have more frequent screening. So this is already coming into action at the moment.

So are there any outputs from it being used?

Yes, we’ve already had women who have had their breast cancers diagnosed earlier by having screening earlier, who have started to take anastrozole. We have a study of 8,000 women which is just about to start in Manchester where we will be offering this risk assessment at their first mammogram and then the decision will be made about those at lower risk, we’ll have a discussion about reducing their screening, and these are women whose risks are below the average woman who is aged 40 who doesn’t get screened, and at the higher end look at the uptake of drugs like anastrozole and of increased screening. We already know from our first PROCAS study which was 58,000 women, that there is extremely high uptake of screening in those we identify at high risk, in the high 80%, and that 97% of those women return for their next screen, which is very significantly more than the standard re-attendance rate.

Do you think this will be available universally?

I would think so, it’s just a matter of it being grasped by the relevant health commissioners to decide this is something we want to do. There are potential savings because, for instance, anastrozole has been shown by NICE to be cost saving to the health service. So for a drug that costs 4p per day you’re halving the risk of future breast cancer. So you save all of that cost of treating those women for breast cancer.