Proteomics promises a revolution in preventative medicine

0:17Hello, welcome to the Naked Scientist podcast. This is the programme that brings you the biggest breakthroughs and talks to the major movers and shakers in the worlds of science technology and medicine i'm chris smith and today in association with proteomics international we're hearing about a new era in medicine where we can tell people not what diseases they've got right now but what ones they will have in 10 years time giving them a chance to turn things around

0:52When I went to medical school, on one of the course documents that they gave out, there was a quote on the cover that said, a healthy person is one who's been inadequately screened. In other words, every single one of us has a skeleton or two lurking in our biochemical closets, which, given enough time, will ultimately lead to the development of a disease. Now if we knew who was going to get what, we could guide patients about lifestyle choices and offer very effective treatments to minimise those risk factors and hopefully kick the disease can down the road sufficiently far that something else becomes a problem first.

1:29So the key is identifying markers that predict the risk and which are present from very early on in a disease course. And if we can spot those, then we can intervene much earlier, well before a person actually becomes ill. One area of medicine where we might be able to make massive strides is in the management of kidney disease. Andrew Frankel is a consultant nephrologist at Imperial College. Well, I've spent most of the last 35 years caring for people with advanced kidney disease, people who reach end-stage kidney failure and lose the function of their kidneys and require either to go on to dialysis or receive a kidney transplant.

2:12But over the last 10 years, my focus has really changed and is now very much on how we might prevent people reaching end-stage kidney failure rather than just treating them for end-stage kidney failure. What's the size of the clinical problem when it comes to kidney disease? How much of the health service's workload and budget does it account for? It is currently the 10th commonest cause of death from any health condition worldwide and it's predicted to go up to the 5th by 2035. And in the UK, up to£7 billion a year is currently spent managing kidney disease. The vast majority on end-stage kidney failure.

2:57We anticipate that unless we do something dramatic, that is going to significantly increase in the next 10 years. And who is this affecting? Who's getting all this kidney disease? Well, the commonest causes are people with diabetes, high blood pressure and cardiovascular disease. accounting for around 10 % of the UK population. And how do we pick it up then? Is it normally people presenting once the horse has bolted, they've already got problems and then they end up in front of you or are there warning signs? We've spent so many years trying to manage this after the horse has bolted. We've recognised people who've got abnormal kidney function or abnormal urinary findings but what we hadn't appreciated is that we are picking people up.

3:47systematically much, much too late in the course of their disease when there is already significant damage to their kidneys and also associated with that damage to their cardiovascular system. So if you knew who those people who do end up in front of you were going to be, say two or three decades ago, do you think you could have changed the course of the disease for them? That really is the hope we now have, because over the last 10 years, there have been fantastic advances in relation to developing drugs and treatments that to slow down and indeed prevent progression of chronic kidney disease.

4:27And if we could just get those treatments to people right in the early stages, not only would we prevent people reaching dialysis or needing a transplant, but we would prevent a significant amount of the damage caused by chronic kidney disease. The problem then becomes, who are those people? Because then you know who to treat. Absolutely. And the current means we have to identify those individuals often rely on tests that pick up the disease when damage is already occurring. And while that is still an improvement on where we were five to ten years ago, what our dream is, is to be able to pick up these individuals before the damage has occurred.

5:11We always jokingly say in my field, which is virology, that we're getting better, but we were very good at telling people what they either recovered from or died from, but not very good at stopping things happening in the first place. Do you think we're now in a position to change that then? Well, I think we have to change it. If we don't really take action, the cost to the NHS and all those millions of people themselves in relation to end-stage kidney failure is going to be huge. And we really have that opportunity. And we can make the need for people to dialyze because of diabetic kidney disease, particularly a preventable disease.

5:52How do you propose to do this? So systematically, there's a lot of work being undertaken across the health service, working now across both primary and secondary care, increasing the screening for kidney disease and increasing the coding of those patients in GP systems so they can be picked up regularly, monitored and have their treatment optimised. The problem is that even with that screening service, we are still picking up people who are already developing damage. We need a marker that tells us who, particularly with the commonest cause of kidney disease such as diabetes, who is going to develop the damage rather than pick up the damage itself.

6:41So if we do have markers that we can use to highlight the patients, do we actually know, though, that we can intervene meaningfully? So although you've speculated what the mechanism of these diseases is, and that by minimising the risk factors, we should be able to change the course of the disease, do we actually know that yet? Yes, I believe we do. And I think the most important thing, though, is engaging in informing and educating the individuals themselves so that they understand what's down the line for them and what can be done to alter that outcome. So there are a significant amount of lifestyle interventions that are important.

7:21But as I've already said, we now have drugs that will slow down and prevent damage occurring. Andrew Frankel, consultant nephrologist, and he also hosts the For Kidneys Sake podcast if you want to hear a bit more from him. So what are the markers that predict a disease and how can we find them? Well this is where Richard Lipscomb, Managing Director of the diagnostic company Proteomics International comes in. I first saw Richard speak at a medical conference in Western Australia about a decade ago. There he blew me away when he showed that if he measured the relative levels of just a clutch of biochemicals, proteins in the bloodstream, he could predict with very high accuracy which patients with diabetes would develop kidney failure and be in need of dialysis, like Andrew was just saying, in the next few years.

8:15The key, of course, is that the process behind a disease leads to subtle changes in our biochemistry, and these are reflected in the different levels of different protein markers. and by considering a number of them together you can greatly improve the diagnostic and predictive accuracy far beyond what you can learn just by for instance reading a person's genetic code as Richard explains. We need to look at proteins because proteins change over time. The genes that we're born with are the genes that we die with but as we grow older then the proteins that we have in our bodies the amounts of them change.

8:51We sometimes start making up of the right protein, we make too much of the wrong protein. And those changes are influenced by environment, by age, infections that we have, or by diseases that we're developing. And so measuring the difference in those protein levels is something that we can detect with the technology that we use at proteomics. If I think about my body, I've and you've got millions of different biomolecules in us washing around in the bloodstream. So how do we know which ones to measure? The reality is we don't in the first instance. The technology that we use is about mapping the entire system.

9:28So in this context, proteomics is the study of every protein in a system in the same way that genomics is the study of every gene in the human body. And so by mapping everything that's there and by comparing people who have a disease to people who do not have a disease, we can look for very subtle changes in that profile and work out which proteins are important for a particular condition. You must have to look at a lot of people then, because everyone's different. There's eight and a half billion people on Earth, and with the exception of a few twins here and there, we're all genetically individual, and therefore our biochemistry is going to be subtly different person to person for every single one of us.

10:09That's a huge set of degrees of freedom. You're right. There's a lot of similarity between people and there's a lot of differences between people. at a population level we can say that there's an average for each protein that's present and by comparing enough people we get into studies of hundreds or thousands of people we can look at the average level of a protein in a sick person versus the average level of a protein in a healthy person and that difference can actually be picked up the technology is really sensitive that we can shine that light on it and understand those subtle differences. And is the approach then that rather than just look at one protein if you look at a constellation of them then the degrees of freedom does go down so rather than just look at say one thing and link one thing to one disease looking at a handful of different proteins that are all affected by that disease in a range of different people you would begin to see a pattern that could be predictive of that condition yeah that's a great way of looking at things that constellation that pattern is what's different that's where we can go now with personalized medicine by just looking at the change in one protein in a body we're very unlikely to be able to tell exactly what is going on but the technology that we have now with a technique like proteomics is we look at all the proteins that are there and we can find that panel of proteins that are linked to a certain disease and that pattern for that panel will tell us what's going on.

11:39Effectively then you're getting at the process of the development of a disease rather than just the manifestations of a disease. If I've already got a disease and we're picking up changes in various chemicals in my body because of that disease then it's good for diagnostics but it's not going to help me prevent not getting that condition. Whereas if you're able to say well this combination of chemical changes tends to crop up in someone who's going to get that disease, that's the process causing those changes, then I presume you can intervene because you can say to that person, this is what's going to happen to you.

12:16Yes, that's exactly it. Changes in process within the body cause the body to make different amounts of proteins. Some proteins go up, some proteins go down. By telling which path the patient's on in terms of where they're heading for disease, we can then say you need this drug for a particular disease. How far ahead do you think you can get of a disease? Because there's this one thing to tell people next year you're going to get X, Y or Z and it's too late. But if you could tell them that 25 years beforehand it might make a huge difference. So where along that line do you think we could get? Yes that's a fascinating area to explore is how far we can go.

12:55With the technology that we have today we're seeing differences three or four years in advance that we can pick up that change at the molecular level, that change in that molecular fingerprint that will tell us where a person is going to be in three or four years time. Whether we can get to a point of picking up that fingerprint 10, 15 years in the distance, time will tell. We're not there yet. I suspect that we will be able to start picking it up in due course. Will you be also able to integrate other sorts of information rather than just the proteins you could have that as one layer but then bring in for instance the genetics on top other clinical factors and then ask a computer to consider all these things together which might bring us further upstream of a person developing a disease it's certainly the case that the body is a very complex mixture of proteins metabolites genes and all of the levels of each of those tell us something about the state of health.

13:53And so there is going to be a pattern there that if we can link it together by using sensible computing and some smart analysis of that data, I would expect us to see a pattern that we can really start to dig into the depths of a disease and where it's coming from. And each of the different molecules in the body is going to add a little bit to that story. So yes, looking at the totality of the system is definitely the way of the future. Richard Lipscomb, Managing Director at Proteomics International.

14:48Naked Scientist podcast with me Chris Smith and today we're looking at the field of proteomics and how it can be used to predict diseases years into the future. Now as Richard Lipscomb was saying earlier, different diseases produce characteristic fingerprint changes in the relative levels of certain protein markers in the bloodstream and if you can spot them you can make very accurate predictions about long-term disease risks. But how do we find the right markers in the first place? Well, put simply, you look at massive numbers of molecules in massive numbers of patients and you try to spot the patterns.

15:25Kirsten Peters leads on this as head of clinical studies at Proteomics International. So what we do is we take somebody's blood and within that blood we know that there are a number of markers or chemicals and what we do with those markers is we break them up into smaller fragments using a range of techniques. And then we run these fragments through specialized machines called mass spectrometers. And these are instruments that can weigh molecules with incredible precision. So each of these fragments of the marker sort of acts like a barcode. And if we measure all of these barcodes in a person, this tells us which markers are present and in what amounts.

16:12And so we can compare the markers in people that we know have disease and people that we know don't have disease and start to get an understanding of differences. So we might have someone that has very mild disease, someone that has moderate disease and someone that has severe disease. And what we do is compare these fingerprints between the different people. And so within the three different groups, we would have lots of people that we would measure. And so we're looking for sort of an average picture of what a mild type of disease might look like compared to severe and also compared to the moderate range.

16:51But people vary by sex. They vary by age. They vary by ethnicity. And they also vary by how bad their disease is. and the disease itself can affect the performance of or the levels of some of these markers, can't it? So how do you weave your way through that minefield? Yes, you hit the nail on the head. The way that we get through that is to use complex statistical modelling. And sometimes this is machine learning. Sometimes it's AI to pull the patterns from the data that no human eye could see. And so this might mean that once we look at the molecular fingerprints, we then overlay information about sex, ethnicity, and we sort of come up with a statistical model that allows for the adjustment of all the different factors that could contribute to this disease.

17:49And what is the ultimate product? Is it a test algorithm that you sell? Or are you aiming to sell a test kit, as it were? Because obviously you're a business, you need to pay for all this very expensive research. So what is the business model? So I think the business model is more we are trying to sell the assay as a whole. And the algorithm, obviously, is secret squirrel stuff. We have worked very hard over a long period of time to come up with this algorithm. And so that would always remain in what I'd call a black box. People can order the test from us and then the blood samples would come to our lab and we run the markers in our lab and then a report would be sent back to the doctor that the doctor can then discuss with the patient.

18:40And so the result is a simple test score that would come back and it would say you have a low, moderate or high risk of having a particular disease. So we've taken an approach of like a traffic light. Green is good. You can go. Amber is you just need to take some caution. And then red is high, like stop. You know, you're a high risk person. And with each of those categories comes an individualised treatment. Where are you now with all this then? What's been achieved and what doors does this now open? So as of now, we have three diagnostic tests that we have actually launched and they're available to the public.

19:27And the three tests are called Promarker D, which is for kidney disease in diabetes. There's Promarka ESO, which is for esophageal cancer. And then there's Promarka Endo, which is for endometriosis. And why are they an improvement over what we have already for those conditions? Yeah, so I might just step into each of them and talk a little bit about why they're so good. So Promarka D, which is the test for predicting in the future. this is a unique test we can predict up to four years in advance somebody is going to develop kidney disease and this is specific to people with type 2 diabetes we do have some preliminary evidence in type 1 diabetes and so we're still collecting data in that space now the problem of kidney disease and diabetes is that it's a silent killer and we have standard tests that patients would have on a routine basis, but we know that these standard tests only identify people once there's significant kidney damage already occurred.

20:41But with proteomics, we've managed to develop a diagnostic that can spot protein changes. So that's those fingerprints I mentioned earlier, years earlier, and this gives doctors time to act before the later stages of disease. The second test is ProMarker ESO for esophageal cancer. And we know that this is a very fast moving cancer that often you only know about once you reach the later stages of the disease. By the time the symptoms appear, it's pretty much too late. So we have identified protein shifts, so changes in those fingerprints that I mentioned. and we have translated this into a diagnostic test where we can identify people early when we can still treat them.

21:30So this test is very important in the space of reflux disease. So there are a large number of people that suffer from reflux and reflux is considered a risk factor for this particular type of cancer. And so currently people at risk of this cancer will undergo endoscopy, which is an invasive technique, very costly and put a huge burden on the healthcare system, not to mention the patients. And so if we could offer them a blood test to rule out cancer, that would make lives a lot easier. The third test is ProMarket Endo. Now this test is for endometriosis. This disease affects one in seven women and girls, is a debilitating disease where women have severe pelvic pain.

22:19Often it takes a very long time to diagnose this test, up to 10 years in most cases. So that's a really long time for women to be suffering. And again, in our lab, we have identified a protein fingerprint of endometriosis that we've turned into a test that could give women answers in weeks instead of years. So it's going to shorten that decade of uncertainty into a clear diagnosis. So together, I think these three tests really show what proteomics can make possible. We can prevent silent disease. We can reduce diagnostic delays and we can catch aggressive cancers early. Really exciting stuff. That was Kirsten Peters.

23:08Now surely the most important aspect of any diagnostic process is the economic argument. But what does that look like? And at the end of the day, how receptive are we all likely to be to future risk warnings of the kind that these tests might be able to offer? Well, I put those points to Smita Sinna, who's a consultant nephrologist at Salford Royal NHS Foundation Trust. So prevention obviously is better than cure, because you might be able to stop people getting these diseases. The challenge we have is how do you actually do prevention? When we think prevention, we think about public health. So that's when we tell people to stop smoking, stop drinking, exercise more.

23:51But there's somehow a way that you go from that to having a disease. And there's somewhere in the middle where we could potentially do better at telling people, if you do these things, then you are less likely to get these diseases. Or if you do more of these things, then you will get a disease. So that might help people make right choices. And we've got a gap at the moment. So Kirsten's talked about these tests. Wouldn't it be nice to have a test that said you're at risk of this condition, you can change things now. And I think that would be enormously helpful. Do you think your patients would listen?

24:26Absolutely. I mean, I get my patients late. You've heard from Andrew how people are diagnosed late. Most of them, when they come to me say I wish I had known earlier and that is a consistent message. So imagine if we gave them that information I think it would make a difference. Even if we give them that information at a late stage they do make changes. It's just much more helpful to make those changes earlier on the disease. This is a programme all about proteomics but the thing that really matters when it comes down to health provision is economics. That's all that people care about is the bottom line really isn't it is the cost so do you see a position for something like this as a way to save money?

25:08Absolutely we know that let's take kidney disease dialysis is really expensive it costs about£35 ,000 per year per person to be on dialysis and it's not a great experience for them they'd much rather not be on it imagine if we could stop them getting to dialysis by identifying their condition earlier identifying their risk and treating them so they never needed it it's obviously cheaper so treating disease early is cheaper than leaving it to the end so I think there's an economic benefit the challenge is finding the money to do the early bit first. Why does the test that we've been discussing one of them that's relevant to you test for what may be coming to your kidneys down the track why does that not exist already why is that hard to do to point the finger and say this person is at risk of this happening to them in about five years?

26:01I mean this podcast summarizes what's happening to science in general doesn't it? Science is accelerating and I think we've seen that in medicine and particularly in kidney disease. We talk about genomics, genetics, proteomics, all these omics but these are all things that tell us more about the human body and its interaction with the environment and our genes. So the science is there, we're now translating that. So making that accessible to people, I guess it's our job as healthcare professionals and the people that make the decisions around what to do with our money, how we use that science and put it to use.

26:38And I think that's the thing that we need to do now, take all that knowledge and then give it to people like me and to patients so that they can make the right choices. And what do you think are the major targets then? There's obviously kidney failure because as you said dialysis is very expensive but what are the other big ones we should be going after kirsten told us that they're looking at gastroesophageal reflux disease they're also looking at endometriosis there must be many other health conditions that you think warrant this kind of advanced warning yeah there are so many things i mean if we look in the uk we you know you hear that we have a challenge with diagnosing cancer late so cancer is obviously an important one but there are also conditions long-term conditions like kidney disease and heart disease that affects so many of us as well as things that affect our day-to-day quality of life indigestion but also mental health issues so i mean you could probably pick any disease and find a way of trying to predict risk so that we give people more information to change things so pretty much everything it's a real opportunity so it gets your seal of approval absolutely i think it'd be nice to have it in my hands so I can use it every day.

27:50I think I would. And I think I would too. Smita Sinna there from Salford Royal NHS Foundation Trust. So from hearing him speak a decade ago to seeing tests for a range of diseases rolling out internationally, it has been very inspiring to follow the pioneering proteomic path that Richard Lipscomb and his colleagues have forged. It is very early days, but this is very much a taste of things to come. We now have technology that lets us shine our molecular microscope on what's going on. It's going to let us develop new diagnostic tests that will really change the way we understand disease and open that door to prevention is better than cure.

28:29Turning disease from something to which we react to something that we can proactively prevent, that must be the future of healthcare. That's it for this episode. Do join us on Friday of course for a look at the latest science news that's hitting the headlines and meanwhile if you enjoy what we do every week and you'd like to help us with our running costs please do visit nakedscientist.com forward slash donate you can make a contribution there we'd also really appreciate reviews on whatever podcasting platforms you use to listen to us and you can follow us for the latest news on linkedin on instagram and on x the naked scientist is supported by rolls-royce i'm chris Smith and from all of us here at the team thanks for listening and until next time goodbye