Julie Williams: Developing Alzheimer's tests

Chris Smith spoke with Julie Williams about promising developments in spotting the signs of Alzheimer's disease...

Chris - So there are lots of different factors happening at once. It's not just that there's one cause, although there might be in some people, I suppose, but you've got lots of different factors all playing a relative contribution. So how can you unpick, then, which ones really matter?

Julie - The secret is to try and use genetics now as a sort of platform to then understand the biology. So we look at individuals that have some of the risk factors that are involved. We are now creating stem cells that reflect the risk in those individuals, we also look at particular cells that we know express a lot of the risk factors, and one of these cells is called a microglia. This is a cell that has housekeeping properties, it has its own little patch of brain that it monitors and keeps clear of toxins or bad tissue and what we know from the genetics is that a lot of the genes that we have found as risk factors are expressed in microglia, some solely in microglia. So that's a big clue that they're playing a significant role in part of the pathway to Alzheimer's disease.

Chris - So does this mean then that chiefly this is a genetic disease, Alzheimer's?

Julie - No. Genes play a role. We look at these on a population basis, on an individual basis. One individual may have a very strong genetic contribution, another individual may have a more environmental contribution. So now we need to get down to precision medicine, a focus where we can identify people that have certain risks, certain susceptibilities, and we may have different treatments or preventions that are focused on these individuals in the future, compared to others.

Chris - How many genes, then, do you now know that strongly influence your likelihood of developing Alzheimer's?

Julie - Well, we think we are about to publish our first genome wide association on early onset. When you add those into the mix, we probably have about a hundred genes that we now know are associated with Alzheimer's disease.

Chris - And so if you took a sample from somebody and you read those hundred genes, how accurate could you be with a test for whether or not a person, that person, is going to get Alzheimer's now?

Julie - Well, we are a bit cleverer than that. What we do is we take those hundred genes and we look at all the other genes that may make a contribution, and we've created an algorithm that probably has thousands of genes involved. From that we can predict, at the extremes, with 90% confidence, whether you will or you won't get Alzheimer's disease. However, the majority of people will be somewhere in between in terms of their risk. Now that's important because, when we do find treatments, we want to identify people at the highest risk so that we can prevent the damage. That's the main, the golden aim, is to get in there before the brain is damaged. I think we can do that now but, in future, the real value of that will come when we have treatments to help people and use that information in a positive way.

Chris - That presumably gives us the 'if' you're going to get it or not, but can you say anything about 'when' yet?

Julie - That's more difficult with a number of factors that influence the common form of Alzheimer's disease. If you have the very rare forms of mutations, yes you can. I think it's going to be more difficult to do that.

Chris - But David Cameron was recently interviewed when we saw some of these new immune modifying drugs were being trialled in people with Alzheimer's disease and he was saying, pretty soon, we're going to have a test which will tell whether a person is going to develop this within the next X number of years. Is that genetic or does that look mainly at the biochemistry of the brain?

Julie - I think that probably would've been a genetic test. We've worked with a company and there is a test out there now. We wanted them to use the best information, so we gave them all our data so that we could best predict those at the extremes. But the problem at the moment is we can't offer people much to prevent that development. But I think very soon we may have more drugs, or maybe even genetic therapies in 5 or 10 years time, that can influence your risk of developing the disease that you can take at an early stage. Things are moving quickly.