Risk and Genomics

Hi, I’m Fran Gale, Head of Science Engagement and Enrichment for Wellcome Connecting Science, and I’m joined Dr. Jon Roberts, who is a Genetic Counsellor and Ethics Researcher for Wellcome Connecting Science. So Jon, tell us about your job. You’re a genetic counsellor. What do you do as a genetic counsellor? In a clinical setting, I work with the NHS, and what I try and do is I support individuals and families as they are making sense of an inherited component to disease. So genetic counsellors work in lots of different settings. I’ll give you two examples. One is cancer genetics.

So someone might have a diagnosis and family history of cancer, and a genetic counsellor’s job is to assess the family history, see if there’s any inherited component, and if there is, if any genetic testing would be warranted or appropriate. Another example of where a genetic counsellor might work is in prenatal genetics. So somebody might have, again, a family history of an inherited condition, like cystic fibrosis, and they may want to use that information to avoid passing that genetic condition onto their children. And again, genetic counsellors can support people as they make those decisions, and help them kind of come to a decision about what they want to do in terms of using that information in a prenatal setting.

You talked about risk. So what is risk in relation to health and disease, and how would you explain that to a student, for example, if they’re trying to understand that? So I think when you’re talking about risk, you need to think about two aspects of risk. So one is an empirical chance or probability. So this normally expressed either as a percentage or a fraction. So you might, for example, say you have a 20% risk of getting prostate cancer over your lifetime, or a two in 10 chance. So you might say something like, two out of 10 people with your risk factors will go on to develop cancer. But that’s just the chance or the probability.

For something to be a risk, it also needs to be a threat, and this is where you have an emotional aspect to it, and also a subjective aspect. So the kind of probability is the kind of objective aspect. The risk is also a subjective aspect. How do I feel about this? So for example, if someone is particularly anxious about someone, that risk might feel big or it might feel small. And to understand risk, you need to understand both the objective chance, but also the emotional aspect of it. How does someone feel about this? What’s their assessment of this event, which can change to whether the risk is high, or low, or even a risk at all.

So somebody might think of something as a risk. Somebody else, they’re not bothered about it. It’s not even a risk. So it’s got those two aspects. So a term some teachers may have heard of is monogenic risk and polygenic risk. Could you explain what those mean? So when you’re talking about monogenic risk, you’re talking about whether a single gene or a single genetic variant that’s causing a trait. So this is the genetics that we’re used to talking about and is often most commonly taught in schools, what you might think of as Mendelian genetics. So the inheritance patterns– dominant, recessive, and X-linked– are all examples where there’s one gene inherited in one of the known inheritance patterns.

Polygenic is where there’s lots of different genetic variants, or genes, involved contributing to a risk factor of a disease. With polygenic risk, often it’s genetic factors combined with lots of other factors, like environmental factors and often chance factors, and lots of different things. The picture is much more complex. So one way of thinking about polygenic risk score is that the polygenic factor is a single piece of a jigsaw. So on its own, it doesn’t make that much sense, but it can be useful to kind of complete part of the bigger picture. Say someone took a genetic test and said they had a 20% risk of developing something like pancreatic cancer. What does that mean?

So there’s a number of ways of thinking about what that risk would be. One is to think about the time frame for that risk. So often when you’re talking about lifetime risk of cancer, you’re talking about that up to age of 80 or so, which means by the age of 80, there’s a 20% chance that you might get a particular form of cancer. One of the ways you might do that in a clinical setting is to break that down. If someone’s kind of particularly young, is you might say, well, actually, there’s also a five year risk and a 10 year risk. So in terms of thinking of what that actually means for an individual.

You can of look at it in those different ways. There’s also ways that you might think about that in terms of different ways of explaining that and framing that. So you might frame it in terms of the risk of getting something, so in this case, a 20% chance of getting cancer, so two in 10 people might get cancer. And the other way is eight out of 10 people, in these situations, won’t get cancer. It can help when people think about it in different ways. And then also, thinking about what that risk factor means in terms of other ways that you might be able to modify that. So this is based on just one kind of simple genetics point of view.

You might say that people with this genetic factor might have a 20% lifetime risk of cancer. You can put that into context by maybe comparing that to what the risk is in the general population. Is it much higher or much lower? But then also ways you might be able to modify that risk in terms of other ways that you might be able to kind of reduce that risk, through sort of environmental, or diet, or lifestyle factors. How do we calculate risk? How is risk calculated in a clinical setting? So there’s lots of different tools you can use. The basics is that you use empiric risks, so data from large populations give you empiric risk.

And then you can also use that with things that they call Bayesian statistics, which is a way of trying to work out how multiple factors can contribute to giving you a risk figure. In the first instance, it depends, really, on what you’re doing, but if we stick with something like cancer genetics, which is a large part of what genetic counsellors do, the first thing you might do is draw up a family tree, because that can give you a good idea about whether or not there’s an increased risk or not. So in that sense, you’re just kind of looking to see, is there things like more cancer than you expect, by chance, in the population?

Because about one in two people will get cancer anyway. So you drop a family tree, you will see some numbers of cancers, but if you see, say, more than average, or if you see lots of linked types of cancers, that might give you an idea there’s some increased risk. You also might look for things like cancers diagnosed at a younger than average age as well. So that’s kind of pre-assessment to kind of say, well, is there some sort of increased risk or not? What you then might want to try and do is try and refine that. Is there a way of kind of putting a number on that? Is there any genetic testing that we can do?

And then you can start to kind of plug in all of those factors to start trying to give people what we call these personalised scores. And what we’re doing is we’re getting better and better using these different factors to give people personalised risk scores, so understanding each of these different factors, their family history, hormonal factors, lifestyle factors, polygenic factors, monogenic factors. So the idea is that you kind of collate all these different factors, then you can kind of give someone sort of more personalised or precise figures, and as we kind of develop the science, we’re getting better and better at that.