ISTAART Relay Podcast – Biofluid-based biomarkers PIA

Adam Smith:

Hello and thank you for listening to our ISTAART PIA relay podcast series brought to you by NIHR Dementia Researcher. ISAART is a professional society and part of the Alzheimer’s Association, representing scientists, physicians, and other dementia professionals active in researching and understanding the causes and treatments of Alzheimer’s Disease and other dementias.

Adam Smith:

In this five part series, we’ve asked members of ISTAART professional interest areas to take turns at interviewing their colleagues and being interviewed themselves. Confused? Don’t worry. It’ll all become clear as the week progresses. We will be releasing one of these podcasts every day in the build-up to the Alzheimer’s Association International Virtual Conference to showcase the work of ISTAART PIAs. Thank you for listening.

Professor David Scott:

Well, hello everyone and thanks for joining us. I’m David Scott. I am an anaesthesiologist based at Saint Vincent’s Hospital in Melbourne, Australia, and I’m a professor at the University of Melbourne as well. I chair the Perioperative Cognition and Delirium professional interest area, but today I’m delighted to be talking with Professor Henrik Zetterberg. Now Henrik’s in Gothenburg now and I’m in Melbourne now, so we’re almost exactly at opposite sides of the planet. Hi, Henrik. Could you tell us a bit about yourself and your PIA?

Professor Henrik Zetterberg:

Hello, Dave. Thanks for having me. My name is Henrik Zetterberg. I’m a Professor of Neurochemistry here in Gothenburg, University of Gothenburg, and also at University College, London, where I run a bio-fluid marker lab at the Dementia Research Institute. To me, it’s been super interesting to see the two different ways of handling the COVID pandemic, because my London lab is completely closed but we are open here in Gothenburg. Trying to do our work the best we can.

Professor David Scott:

That’s very fortunate for you.

Professor Henrik Zetterberg:

It’s been a special, a very special spring this… Yeah, so I chair the bio-fluid based biomarker PIA, and we are focusing… Actually the PIA started with a big focus on Cerebrospinal fluid measurements, but recent developments in this field with ultra-sensitive measurement techniques have really made it possible to transfer some of the CSF biomarkers to blood tests. We renamed the PIA to make sure that these developments were covered in the work we do, and then we have been able to work across the different APS much more. Not the least with your PIA Dave, looking at brain effects of anaesthesia and such things.

Professor David Scott:

So let’s start with… I guess, CSF, which is the bio-fluid that you presumably started with. Particularly with your link with Alzheimer’s disease. So is CSF old fashioned now, do we need to bother about collecting CSF samples? Or a CSF still the gold standard for bio fluid biomarkers and what can we learn from CSF biomarkers?

Professor Henrik Zetterberg:

That’s a great question, and of course, I love CSF and I’m very CSF biased. It’s a beautiful body fluid, the best body fluid one could think of… Clear, nice, easy to work with. It’s very easy to measure biomarkers in cerebrospinal fluid because of its proximity to the brain tissue. So the idea is that CSF communicates rather freely with the brain interstitial fluid and then biochemical changes in the brain should be directly reflected in the CSF. But of course, that is an oversimplification because CSF is to large extent the plasma filtrate and the minority of the volume comes from the interstitial fluid in the brain. It’s around 70 to 80% comes from plasma, but the remaining percentages are brain derived and it has been easier to measure the CNS molecules in CSF because the concentrations are often around 50 to 200 fold higher in CSF of the brain derived molecules.

Professor Henrik Zetterberg:

So, I think that the traditional reason for why CSF has been a preferred bio-fluid to study brain diseases. But then there have been extremely nice developments in the measurement technologies. Now we have techniques like single-molecule RA, like [inaudible 00:04:57] and also, proximity legation assets, and other amplified… Other assets where you can amplify the signal of the measurement. Better mass spectrometry-based methods also. And then we have been able to measure many of these molecules in blood.

Professor David Scott:

So we’ve… It’s been quite a few years since we heard about the trial, if you like, of Abeta and tau and p-Tau and how in CSF, there was this fingerprint almost, of potentially an indication for the presence of Alzheimer’s disease, or even the severity of Alzheimer’s disease. Does that still hold true? Or do we have better biomarkers now?

Professor Henrik Zetterberg:

Yep. This trial still… Is still highly relevant. And if we start with the amyloid pathology, of course, you can measure it, the amyloid PET, and then you get an anatomic information. But for amyloid pathology, that anatomic information has turned out not to be very important, at least not in the symptomatic disease stages. The best fluid biomarker for amyloid pathology is the ratio of 42-amino acid long beta load to 40-amino acid long beta load. And that has… It makes very nice concordance with amyloid PET. So from my perspective, these are almost interchangeable. Trying to transfer the amyloid test to blood test has been challenging, but recently there are now a number of publications indicating that the [inaudible 00:06:31] may be the 42 to 40 ratio actually pinpoints amyloid pathologist as determined by amyloid PET or the CSF Abeta-42 ratio, rather well. But there is a problem with amyloid as a molecule, because it is also expressed in blood platelets and release from blood platelets it’s expressed in the liver and some other non CNS tissues.

Professor Henrik Zetterberg:

So in CSF, there is a 50% reduction in ratio, but in plasma. The reduction, if you’re amyloid positive, but in plasma, the reduction is 14, 15, 16%, and that makes it hard to reliably measure such a small difference. And then the tests must be very precise, but the mass spec based tests that have been developed in Japan by Nakamura and also in st. Louis by the Randall Bateman team. They look really promising, although it will be a little bit problematic. So we are now having simple blood tests for amyloid pathology. We can do them in CSF, and we have amyloid PET. And depending on where you are in the world and how your study is designed, you could use these different markers. And perhaps you could also use them in sequential if needed to increase the specificity of the test.

Professor David Scott:

So you’re certainly drawing the connection between your work and the work of your PIA with Alzheimer’s disease. And I mean this… The possibility of a blood test has always been the Holy grail hasn’t it? For determining whether someone has an illness or not, and how severe that illness is. Perhaps you should tell us a bit more about these highly sensitive, small molecule assets.

Professor Henrik Zetterberg:

I could just also briefly touch base on the Tau part of this spectrum, and also mentioned neurodegeneration rapidly. So for example with Tau we have, of course, big advantages with Tau PET imaging, because then we get anatomic information. And that really seems to correlate with clinical expression of the disease. So the brain regions in which Tau accumulates will reveal four types of symptoms there might be in. And it also correlates strongly with neurodegeneration. The fluid tests for Tau pathology are not… Do not give this anatomic information, which is a drawback. But CSF phosphor tau is still a reliable test to detect the tau phosphorylation that sort of precedes a tangled formation and the development of neuritic dendrites. And this test has also been made into a blood test with an ultra-sensitive method, which is… Which I will touch upon shortly regarding how that works.

Professor Henrik Zetterberg:

For neurodegeneration we have magnetic resonance imaging of the brain as the gold standard method and then also, of course, FDG-PET. But there… We have two biomarkers that have been discussed a little bit in regards to what information they might bring. We have CSF total tau, which traditional has been regarded as a neurodegeneration marker. Now I, and many other researchers think that neurons exposed to amyloid will respond to that pathology by phosphorylating and secreting both total tau and phospho-tau. So it’s not really neurodegeneration, the total tau measure, but perhaps something that will predict neurodegeneration. The best fluid biomarker for neurodegeneration, in general, is neurofilament light, which is an orphan drug zonal protein that leaks out from injured axons. So total tau surprisingly AD specific as a biomarker whereas neurofilament light is more general neurodegeneration marker.

Professor Henrik Zetterberg:

So that is how we try to use them now and both phosphorylated tau and neurofilament light have been made into blood tests that correlate with CSF levels and also with the imaging evidence of tau pathology and neurodegeneration. And what has happened is that the traditional immunoassays, where you capture the target analyte between two antibodies. A capture antibody and the detector antibody, which is labelled by some means. Those assets have now been transferred onto a single molecule area where you have the capture antibody to magnetic beads, and then you pull down magnetic beads with the analyte into small microbials. And one bead will fit into one microbial. And then you will be able in the instrument to add the substrate of the enzyme with which the detector antibodies labelled. And then if there is analyte present, there will be light emission. And then this will happen on almost like a Chinese chess board like grid, where you image… Emission of light.

Professor Henrik Zetterberg:

And then you can count molecules one by one. So if the fluid you are measuring on is diluted so that most microbial will not emit light. You will know that those, the two contains one analyte. If they contain two analytes, they will emit me twice as much light. If they contain three analytes, they will emit three times as much light. So… And this has made it possible to have various sensitive assets with low limit of quantification and also quite nice dynamic range because you can also capture than the analogous signal. So you can also measure high concentrations. This principle is called single-molecule counting or single-molecule count. You have this digital component to the asset and there are similar techniques that work along similar lines of… And they sort of all can achieve this type of ultra-sensitivity.

Professor David Scott:

You’re describing a process with incredible sensitivity and clearly moving into very exciting areas. And it was very reassuring. I’m sure for the neuroimaging here, that there’s still some value left with doing some structural imaging of the brain. That’s excellent and very reassuring. Let’s just talk a little bit about the PIA itself, Henry. Who forms your PIA? What type of people are in your PIA? What type of… Are there scientists? Are there clinicians? How’s that work?

Professor Henrik Zetterberg:

It’s a very mixed composition of the PIA. There are researchers. Many of these researchers are biomarker researchers who work in a focused manner on improving the analytical techniques to measure these molecules. So it’s rather a sort of chemical to some extent. Then we have biotech people and people from the diagnostic industry. So it’s a rather industry intense PIA and then we have…

Professor David Scott:

Come out and talk to us. We understand what you’re saying. It makes sense. So that’s a credit to you and your members.

Professor Henrik Zetterberg:

Then another category is, of course, clinical cohort owners who want to keep up with the latest developments in fluid biomarkers. I mean, many epidemiologists and genetic researchers are now starting to use the blood-based biomarkers as endophenotypes to discover risk factors, not just for the clinical phenotype, but also for a molecular pathology. Or a polygenic risk score, for example, for tau pathology and not just the clinical phenotype. So that type of work is ongoing, quite intensively now.

Professor David Scott:

So are these small molecule essays and the highly sensitive essays, are they ready for clinical application in research and epidemiology research or clinical research? And are they… How close are they to clinical use?

Professor Henrik Zetterberg:

Actually, two weeks ago we started to measure neurofilament light in plasma here in Gothenburg, in our clinic laboratory practice. So now we use it as a general marker of neural injury and we get samples, for example, from primary health care physicians. Some people sent a few days ago from a patient who was around… Well in her twenties and she complained about numbness to her right hand. And she told the physician that she had a little bit of numbness in her left foot a while ago. And this she wanted to check if this was perhaps indicative of multiple sclerosis. So it’s… But one could also think about persons seeking medical advice because of memory problems and then one would use neurofilament light. This is a rather sensitive test for neurodegeneration. And then if that is positive and could perhaps then speed up the referral of the patient to a memory clinic for more advanced examination. We hope to do this for plasma forceful tau so…

Professor David Scott:

So you’re not putting the clinicians out of work just yet. That’s excellent.

Professor Henrik Zetterberg:

Not actually they… Of course, there is a worry that this will generate more questions than answers and I mean that can happen with new tests. But I think we… I think that those… These types of developments will teach us a lot, but of course, one has to be careful discussing with the patient if he or she really wants new tests to be done because… So that the patient is feeling comfortable with what such a result might lead to. Of course, we hope in this PIA that there will be disease-modifying treatments for perhaps for several neurodegenerative diseases in the future. And then this type of work might guide treatment advice in perhaps it could put also speed up how fast patients could get access to different treatments. At least if the phosphor tau and the beta the tests become standardised for clinical laboratory practice use. So then… We think that this might be… Perhaps this would be the most important work for our PIA in the upcoming years. How to make the assets fit for purpose in clinical laboratory practice. Help with what… How to interpret the biomarker results and perhaps also help out to communicate the results with the patients because many of these biomarkers change so early in the disease process.

Professor Henrik Zetterberg:

And of course, we might do patients a disfavour if someone has slight memory problems and sought medical advice 10 years ago, perhaps he or she would be reassured that this commonly happens with ageing a little bit, and it doesn’t seem to be that bad, and the patient would be left without a diagnosis. Now it might be that the patient comes and gets an early Alzheimer’s diagnosis perhaps 10 years earlier than what would have happened 10 years ago. And then we have to be extremely careful with redefining or re-branding what Alzheimer’s disease is and that you can live with this disease and function quite well over an extended time period. I mean the worst thing that could happen with such a biomarker support diagnosis, if this becomes a reality in the future in the clinics, is that the patients get an Alzheimer’s disease diagnosis and don’t think the things that he will get… He or she will be like perhaps an old relative with Alzheimer’s disease that he or she has met.

Professor David Scott:

There’s certainly many significant issues with respect to diagnosis, just not the number on a lab report. That’s absolutely correct. The person really needs to be considered. Just briefly, you’ve also been doing work in some acute changes as well, I understand. Delirium obviously comes to mind. You’ve talked about neurodegenerative diseases, but where does delirium sit in all of this with respect to your work?

Professor Henrik Zetterberg:

I think this is so exciting and so the way we have interacted with your PIA, for example, the hypothesis have been… There have been a number of hypothesis that need to be addressed and they need to be tested further, of course, and that is the vulnerable brain concept. That if you have a brain with a clinical, a silent pathologist having accumulated, and then you are challenged by having to go into intensive care or doing surgery under general anaesthesia or something else stressful that would give you an increased risk of developing… Of getting confused by the whole thing and getting… Developing delirium. That’s one thing is this could silent, for example, alternate pathology or neurofilament light in the upper range before the procedure would have predict delirium. That’s one thing, the other question is if delirium is injuring the brain. So if you have a prolonged time of delirium, will that injure the brain and cause harm by itself, and that we can now start to play around with a little bit with these blood biomarkers.

Professor David Scott:

It’s a bit of a chicken and egg situation isn’t it?

Professor Henrik Zetterberg:

Yeah, exactly.

Professor David Scott:

Is it both going on? Do you think? [crosstalk 00:20:25]`

Professor Henrik Zetterberg:

It could be that type of process that… And I think the data that we start seeing published now, they are well mildly supportive of both concepts, I think. What would you say, Dave?

Professor David Scott:

You’re not supposed to be asking me questions yet. I think I don’t like the term, but I think delirium is a bit like brain failure, like heart failure to parallel it. So a sick heart will fail, but then the episode of failure actually damages the heart more. So I think they sort of combined to each other.

Professor Henrik Zetterberg:

Yeah, that makes sense.

Professor David Scott:

Henrik, If I could move to just… The area of introducing your researchers into our fields, in your field in particular. What do you think early career researchers who are listening can do to become involved if they’re suddenly excited by bio-fluid biomarkers? What should they do?

Professor Henrik Zetterberg:

You know, I think that there are… I think there is… But now this is my bias. I have to tell you upfront, but I think learning the analytical techniques. Exploring new technologists first-hand. Learning things that are really cutting edge and learning them in practice indeed, to really learn the biomarker measurement craft. That is my advice. And of course, and this is my other bias. I am not that good at statistics and especially not bioinformatics. And that complicate, I really have not been able to keep up with that field. But when I meet young, mostly people who are skilled at this, I think there is a lot to learn to do here. So these are my two things. Either you learn the basic chemistry and the analytical technologies in great detail, and really make sure to keep up with the field in terms of the cutting edge technologies to make… Because I think we will… We can measure total concentrations of certain protein, but it has thousands over at least hundreds of isoforms and different notifications. Perhaps one will find interesting things in that. But the other thing is to handle complex datasets with advanced statistics. I could mentor the first part. I could never mentor the latter. I think these are the two things to me.

Professor David Scott:

Your enthusiasm comes through, and I think that’s another whole element of mentoring that people…

Professor Henrik Zetterberg:

Can I just too rapidly comment on a thing I forgot to comment on Dave, and you brought it up in the beginning. Is there a need to do CSF?

Professor David Scott:

Yes.

Professor Henrik Zetterberg:

There is from my perspective less need, if you’re interested in brain derived molecules that seem to… I mean, they appear in the blood and if you have a sensitive enough asset, that’s fine. It could be that enzymes proteases and sort, it could mess up the biomarker in the blood. So the blood is worse as a bio-fluid than CSF, but I think that is solvable.

Professor Henrik Zetterberg:

The part where I really think CSF is needed is if you’re interested in inflammatory processes or… And actually also to a large extent synaptic homeostasis and endosomal vesicle recycling such things, lysosomal proteins. Biomarkers that are generally expressed across tissues because then the blood levels will be influenced more by what is released from example immune cells than what it is from the brain or the microglia in the brain. So there, I think if you do a CSF sampling, you will have a greater chance of seeing something microglia or astrocytic related, than if you look in the blood. So that’s the inflammation, microglia activation, endosomal vesicle recycling, lysosomal function, such things that I think you would still gather additional knowledge by doing CSF examinations.

Professor David Scott:

That’s so what you’re saying in one way is its very hard to disentangle peripheral information from central information unless you sample centrally.

Professor Henrik Zetterberg:

Exactly.

Professor David Scott:

So, I mean, that’s again is there’s a whole other area of exciting research, which can be done. So you… We… I guess we’re all disappointed that we won’t all be face to face at the AAIC this year. Is your PIA going to be presenting nonetheless?

Professor Henrik Zetterberg:

Yep, we are. I’m a bit uncertain exactly how it will be arranged now, but we will have presentations from young researchers like we had last year. I’m not sure exactly which date that will happen, but that was so fun last year and so we decided to repeat it exactly like we did, but over zoom then. And then we… This year we will also have a prize to the highest-ranking abstract, which I think that cross be good encouragement to have that type of… But just being selected to present a prize in itself. But we try to… We will try to use the PIA activities to help networking and… I mean, that’s the main challenge now that we can’t meet in… And there are many people seeking, I mean, Postdoctoral jobs and so. And new projects, new collaborations, all those things have to happen. And it’s good because it’s not just the junior people needing to interact with senior group leaders, senior group leaders need to find excellent people to the team. So, we are all in the same boat and we will try to use the PIA to help with such interactions.

Professor David Scott:

Thanks very much Henrik and I guess you’ve emphasised the underlying principle of the PIA, isn’t it? Which is networking and sharing of ideas and forming collaborations. Which your PIA has done very well and we’ve benefited from that as well. So thank you very much, Henrik Zetterberg for joining us today.

Adam Smith:

Thank you so much for listening. You can find details and profiles on today’s panellists and information on how to become involved in ISTAART on our website dementiaresearcher.nihr.ac.uk, and at alz.org/istaart.

Adam Smith:

We’ll be back tomorrow with the next recording of in our ISTAART PIA relay podcast series. Finally, please remember to subscribe, like and review our podcast on iTunes, Spotify, and all the other places where you find your podcasts. Thank you.

END