The Evolving Role of Biomarkers in Alzheimer’s Evaluation

QUEST WEBINAR TRANSCRIPT:

The Evolving Role of Biomarkers in Alzheimer’s Evaluation

Presenters:

• Dr. Michael Racke, Medical Director for Neurology, Quest Diagnostics
• Dr. Matthew Stroh, our Medical Science Liaison for Neurology, Quest Diagnostics 

Cuttic, Natalie M   0:06
Hi everyone and welcome. We are so glad that you could join us today for our webinar, The Evolving Role of Biomarkers in Alzheimer's Evaluation. My name is Natalie Cuttic and I'm part of the neurology marketing team here at Quest Diagnostics and I'll be your host. Host for today's session.

We're thrilled to be joined by our expert speakers, also from Quest Dr Matthew Stroh, our Medical Science Liaison for Neurology, as well as Dr Racke, our Medical Director for Neurology. Before we get going, just a quick few housekeeping notes.

Today's session is being recorded and it will be available to view on demand through our Clinical Education Center. A link to the recording will be sent out shortly, so if you'd like to revisit any of the content or share it with a colleague, you'll be able to do that when the webinar recording link is available.

Also, as you notice, we're on Microsoft Teams. The Q&A feature is already live. Please feel free to submit your questions at any time during the presentation. We'll be keeping track of them, and at the end of today's program, we've reserved about 15 minutes for Q&A, where I'll pose as many of your questions as possible to our speakers. So with that, let's get started. I'll now hand things over to our first speaker, Dr Racke.

 1:31
Thank you, Natalie.

Racke, Michael K   1:41
Matt and I are both employees of Quest Diagnostics. As such, we've disclosed that for you today.
Today we're going to be looking into the evolution of the Alzheimer's diagnostic landscape. We're also going to review some of the latest criteria and guidelines regarding Alzheimer's testing and staging, and we're also going to discuss some of the implications regarding the use of blood based biomarkers in the clinical care pathway as it exists today and as we think that it's going to evolve in the future. As many of you know, Alzheimer's disease is one of several dementias that can occur in patients.

Dementias are typically neurodegenerative disorders that result in impairment of cognition, and they were initially characterized based on their clinical features. Interestingly, as people began to see pathology. They began to see that some of the pathologies coexisted.

Matt, can you talk a little bit about mixed dementias?

Stroh, Matthew A   3:04
Yeah, certainly. Yeah. As you, as you mentioned, you know, dementia is kind of an overarching term for, you know, a number of different symptoms. And you know, Alzheimer's disease is the most common form of dementia, you know, with 60% of 60 to 70% of dementia cases being actually due to Alzheimer's disease pathology.

But as you mentioned, there are other types and they are, you know, ideally, as you can see here on this on this nice infographic, they're kind of put into these nice little buckets, right? We have Alzheimer's disease, vascular dementia, you know, frontotemporal dementia, Lewy body dementia and then other dementias that include, you know, those that are like CJD or even Korsakoff syndrome. But you know, in reality, the nice little buckets are not what we actually end up seeing both in the clinic.

But also when it comes to the actual pathology and pathophysiology of the diseases. So go ahead, Mike.

Racke, Michael K   4:08
So I think you bring up a good point, Matt, and that is that in the real world, you know, it's not necessarily that we categorize disorders based on their pathology necessarily. We initially characterize them based on their clinical features, but I think as this slide indicates, there's a lot of overlap with some of these disorders. And the other thing that's really important is that there is, you know, anatomic localization based on some of these disorders. Can you talk a little bit about how we use biomarkers And the localization to help us in a diagnosis.

Stroh, Matthew A   4:53
Yeah, certainly. So as you mentioned, you know the original, you know the original approach diseases like Alzheimer's disease was more of a clinical pathologic approach, right. So clinical presentation and then post-mortem examination to confirm pathology.

But you know, as you can see here in these in these infographics, there are a number of diseases that have mixed and overlapping pathology and pathophysiology, which makes, you know, evaluating them in the clinic.
You know, maybe potentially a little difficult, right? Because they can have different presentations or even similar presentations, but have different pathologies, right, based on where they are in the brain. And this can lead to, you know, obviously misdiagnosis and as we enter an era of disease modifying therapy and more precision and personalized medicine, you know we're developing drugs that are really looking to target very, very specific pathologies.

So you know the clinical diagnosis is only part of the story that clinical diagnosis and clinical evaluation and presentation is extremely important, but it also is going to be composed of not only looking at, you know, the behavior and the behavioral and clinical changes, the neurocognitive changes, but also those changes that you can see in the biomarkers as well as anatomical changes.

Racke, Michael K   6:25
So I think you bring up a good point is that, you know, when I was a resident, clearly if somebody we thought somebody had Alzheimer's disease, the IT was really based on the clinical characteristics. And now I think because we have sophisticated imaging with MRI that we have a different picture in terms of how those contribute. Can you talk a little bit about?

How, in addition to the clinical picture, we look at other aspects in defining dementia.

Stroh, Matthew A   7:02
Yeah, certainly. You know as you mentioned like with overlapping presentation in the clinic, you know you can have, you can have you know individuals that have Alzheimer's disease with classic amyloid and tau tangles and pathology, but you can also have patients that present.

Like they have Alzheimer's disease, you know, earlier on in the beginning stages of the disease, but they ultimately have something called, you know, the behavioral variant of frontotemporal dementia, which generally is not amyloid. It's not accompanied by amyloid pathology, right? So when we're looking at the clinical presentation, we also have to ensure that we're understanding, you know, not only in where in the brain, how it's, you know, how it's affecting their behavior, but also what is really causing this.

And you know the complex and mixed dementia pathology is one that we end up seeing later, especially later on in life when you know when individuals you know get into their 80s and 90s in those decades, you know we end up seeing on post-mortem examination it's not uncommon to see, you know, very mixed pathologies, you know, with maybe a patient has Alzheimer's disease, but also ends up having things like, you know, Lewy bodies or, you know, other pathologies.

Racke, Michael K   8:32
I think you bring up a really good point and that is before when we were looking at what the clinical phenotype was, I hate to say it, it almost wasn't relevant. But now when you say have specific therapies targeting specific aspects of pathophysical physiology. Then all of a sudden the biological pathways all of a sudden becomes almost the most important thing, because that's what's going to help you select a treatment.

Stroh, Matthew A   9:03
Absolutely, yeah.

Racke, Michael K   9:07
So, you know, we were talking a little bit about the ways that you diagnose Alzheimer's disease. And I can tell you when I was a resident back then, if patients died in the hospital, they often went to autopsy and for somebody who had an underlying the way we really confirmed Alzheimer's disease was by a postmortem examination, identifying those plaques and tangles in those patients at autopsy.

Subsequently, we learned that there were ways to diagnose Alzheimer's disease when they were alive. Can you talk us a little bit about some of those methods, Matt?

Stroh, Matthew A   9:50
Yeah, yeah. So as you mentioned, you know, post-mortem analysis is, you know, it's the kind of the gold standard, but we have to separate these into, you know, post-mortem and ante-mortem. Obviously post-mortem is not ideal for, you know, those patients that we want to end up treating with drugs that we've developed in clinical trials, which requires patients to be identified obviously while they're still alive.

So you know, postmortem is, you know, it's the canonical classical way of confirming dementia. But as technology has developed and as our interest in really developing these targeted therapies has grown.

We've developed more antemortem ways to evaluate pathology in the brain. So for example in Alzheimer's disease, you know specifically there is the development of the positron emission tomography or PET scans with the first radio tracer to be approved being Pittsburgh B compound. I believe that there are three now that have been FDA approved for evaluating amyloid deposition in the brain. Of course there are also Tau tracers that are being developed and approved for evaluating Tau pathology.

So you know the advent and development of PET scans has been you know for a boon for research and for our ability to identify patients early and you know earlier on in the disease. But in addition to that, we've also you know been interested in evaluating individuals with, you know, fluid-based biomarkers and there are a number of reasons for this. The first being obviously PET scans, you know they're costly, right? But they also expose the patient to radiation and the access can be, you know, significantly limited.

So, you know, there was a move to really develop fluid-based biomarkers with the CSF being the low-hanging fruit, seeing as the CNS is a really privileged system. You know, the blood brain barrier keeps the CSF separated from the rest of, you know, the bodily fluids.

And this CSF is a really good source for us to be able to evaluate, you know, what is, what is the underlying brain health look like when it comes to these biomarkers for Alzheimer's disease specifically, you know, we're looking at amyloid beta 42, amyloid beta 40 and Tau pathology, specifically phospho-taus.

Racke, Michael K   12:32
So I think that that you bring up a good point and that is now all of a sudden we have access when we think about plaques and tangles in the brain, you know, we know that the plaques are caused by beta amyloid and the neurofibrillary tangles are caused by the aggregation of hyperphosphate.

How does that change in terms of our insight into how the disease course changes over time? And I guess you brought up about the cost. What's the difference now when you look at the blood-based biomarkers versus spinal taps versus PET scans?

Stroh, Matthew A   13:12
Yeah, certainly. So, you know, Alzheimer's disease specifically and we know this about, you know, other types of dementia, but you know, we're really focusing on Alzheimer's disease today. You know, Alzheimer's disease is, it's an insidious course of disease, right? It is one that we, you can start seeing pathology change in the brain as early as 10 to 20 years in advance of the disease actually resulting in clinical onset and symptoms. So you know, if we think about that, then you know individuals who may be in their 40s or 50s could actually already be depositing amyloid and have the phospho-tau changes that we see early on in the entorhinal cortex and eventually spreading to the neocortex and throughout the brain reliant upon amyloid. So you know early identification is really important with these diseases.

Now you know as we mentioned the cost of blood based or of PET biomarkers is a limiting factor and the access is a limiting factor. You know CSF is significantly more cheaper but it is also still limiting for a number of different reasons, right? So CSF is, you know, more difficult. It requires specialized training outside of a phlebotomist, you know, to be able to actually perform the spinal tap. It is more costly than, say, blood-based biomarkers. And we've also found that of course patients don't necessarily want to sit through a spinal tap if they if they don't have to, right. So this kind of led us to looking at whether or not we could evaluate and look at these blood-based biomarkers or these biomarkers in the blood, which led to the advent of blood-based biomarkers.

Racke, Michael K   15:10
So you brought up a good point and that is that there's been this evolution from PET scan to CSF testing and now to blood based biomarkers. I get asked all the time about the sensitivity and specificity of these tests. Can you tell us a little bit about some of the terminology that's used in telling us about the performance of these tests?

Stroh, Matthew A   15:36
Certainly, yeah. So you know, sensitivity and specificity is something that, you know, all clinicians are really familiar with and it's the number one question we get, you know, when we're talking about these tests in the field. And the sensitivity and specificity is obviously incredibly important, but it's also something we want to focus on here in that the interpretation can change based on a either single cut point or dual cut point test, which we'll talk about a little bit later. So for the sensitivity, obviously just a quick overview and recap. We take a group of individuals that are known to have the condition and we really look to see how well a test.

You can identify true positives within that cohort and avoid the false negatives as well as with specificity. We're really interested in identifying those patients that are truly negative in patients that or individuals that are known to not have the condition and of course you know, sensitivity and specificity, they're not dependent upon how prevalent the actual disease or the predictor or the comparator are in the patient population for our in our case, you know, amyloid pet positivity.

But you know, positive predictive value, negative predictive value, these are dependent upon and can change with the prevalence in the patient population or the cohort. So it's incredibly important to think about the prevalence of, you know, amyloid positivity and you know, think about this in the context of clinical suspicion of of AD depending on whether or not you're in a memory clinic setting or whether or not you're, you know, a primary care physician.

Racke, Michael K   17:22
I've also heard the terminology of like ROC or area under the curve. Can you tell us a little bit about that?

Stroh, Matthew A   17:32
Yeah. So again, you know, a brief overview, you know, with ROCAUCS, the ROC ROC curve, as you can see in this, this top graph here, just a quick orientation, you know, basically it's, you know, it's an analytical method that's used to evaluate the performance of binary diagnostic classification methods.

And if you look at that dotted line right there in the middle, that kind of splits the box. This is the equivalent of essentially a coin flip. So if a physician sees a test that has an ROC that is pretty close to that line, you know the actual probability or performance of the test in being able to differentiate and provide good valuable information to the clinician is much lower, right? So the clinician actually may be just as well to pull out a coin and flip it in that you know they have an equal probability of being able to determine the status of that patient.

Now if we talk about AUC, this is the area under the curve. And again, this is a way to evaluate the accuracy of a diagnostic test. Now what's important to remember, of course, is that you know AUC's, they can actually change. Or you know they can be the same depending on the different types of tests. But if there is an AUC for a test, the same test can have the same AUC but have different sensitivities and specificities and ultimately be able to, you know, rule out or rule in differently.

Racke, Michael K   19:07
So can you, uh, tell me what's going on here in terms of AUC?

Stroh, Matthew A   19:12
Yeah, I love this little animation. This is a really, really cool way of being able to understand a UCS and again talk about how sensitivity and specificity within a group of patients that are known to have the condition. Let's say this is the group in red, the distribution in the red and then the group of patients that are that are actually known to not have the condition in the blue, right. So as we are able to differentiate based on the predictor and be able to separate those two groups based on the predictor in the test, you see that the AUC actually approaches one with you know an AUC.

Of one being a test that can perfectly differentiate those patients that both do and do not have the disease.

Racke, Michael K   19:58
Right. And so I think it's important to point out we almost never see an AUC of one, but it often is helpful to compare amongst tests what the different AUC is.

Stroh, Matthew A   20:13
Absolutely. Yeah. Especially, yeah, especially in this, in this space, you know, with these complex neurodegenerative diseases, these spectrum disorders with, you know, complex and heterogeneous pathologies and presentations, it's exceedingly rare to see an AUC of one.

Racke, Michael K   20:13
So now, yeah, so now we're seeing utilizing different cut points. What can you tell us about this?

Stroh, Matthew A   20:38
Yeah. Again, this is a really cool animation that kind of shows us where a normal test may have the predictor basically leads to an overlap between those patients that are known to have the condition and not have the condition. And we want to demonstrate here that you can see that the AUC does change, but that you can exchange between the true positive rate and the false positive rate and ultimately identify an optimized cut point that gives you know the maximum true positive rate and the maximum false positive rate.

Racke, Michael K   21:15
And so here it's showing that, for example, if you optimize that, that's the point that best differentiates between those two.

Stroh, Matthew A   21:24
Correct.

Racke, Michael K   21:25
Here, can you tell us now? Now we're starting to get into looking at beta amyloid 4240 and how a clinician might use that in combination with a cognitive diagnosis. Can you tell me what's going on here?

Stroh, Matthew A   21:42
Yeah, absolutely. So you know, our team was interested in evaluating the 40 to 40 ratio, which is a good indicator of amyloid deposition and risk for Alzheimer's disease.

And you know, it's obviously previously developed in the CSF and we were interested in looking at it in the plasma as well as other groups are. And you know, so we took a patient group of 250 individuals with healthy controls, those that have MCI, which is mild cognitive impairment and those that actually have Alzheimer's disease and these patients were really well characterized cohort that we knew. You know all of their, you know, their demographic information, we knew their cognitive scores.

We also knew their amyloid PET status which was ultimately our competitor and you know the thing that we look to be able to predict with these plasma based 4240 biomarkers. So as you can see over here on the left, this is the ADRC healthy control group.

With the amyloid beta PET negative individuals mostly sitting above that that cut off line of .16 which we were just talking about optimized cut offs and you know the vast majority of those patients in the healthy control group sit above that, that .16 line indicating that they're lower risk and you can see that they also have been confirmed to be amyloid PET negative. Now with the mid-group you can see a much wider spread with you know a mix of obviously both amyloid PET beta PET negative and also amyloid beta PET positive.

But the optimized cut point did a very good job of being able to differentiate and you know have the majority of pet negative again above that line and the majority of pet positive below that line. Now you can see here that there are obviously some amyloid pet negative individuals that sit below that line in the lower risk category or a higher risk category, but they are, you know, their amyloid pet status is known to be negative at the time. We'll talk about that here in a few. And then of course the ADRC AD group. Those individuals were amyloid PET, not a positive, and they, you know, set well below with the majority below that line.

Racke, Michael K   24:20
And I think it's important to recognize that by including the healthy controls that often you'll see studies that do that to sort of enhance the performance characteristics of their test, whereas what is typically asked for in guidelines is what's in the intended use cohort. In other words, if you have somebody with cognitive impairment, what's that? You know, what does that show in terms of testing?

Now we talked before, you talked before about sensitivity and specificity. Now what I would like to do is apply that to some real data with regard to PET scanning and beta amyloid.

Stroh, Matthew A   25:02
Yeah, so this is, this is a really great representation of again the data that we just showed. You know it's the same group, same cohort with the ratio of plasma AB to 4240 over there on the left and just to Orient ourselves here.

You know down in the bottom right these are individuals that were that were amyloid pet positive at the cutoff of 1.4 and greater being amyloid pet positive and then having a detect plasma ratio of 40 to 40 below .1 .6 which indicated that they're a higher risk. These are the true positives there in the bottom right and of course of over here in the top left, these are the true negatives.

Now if we look at the bottom left here, these are those individuals that are false positive. They have an 80 detect score that shows that they are at high risk, but their amyloid PET status is actually showing that they are they're negative on Amyloid PET. Now you can see that there are obviously a number of people that get very close to the cut off, but this is where I was talking about that overlap between groups and optimizing cut points.

It's also important to note that you know with a number of these individuals that are false positives, you know the vast majority said in the MCI group and the MCI group, you know these are individuals that ultimately will likely convert to Alzheimer's disease and they're actually in the process of depositing amyloid and have not actually shown up to be positive on PET yet.

Now the top right quadrant here is the most obvious, right? This is where we see the false negatives, which indicates that this test is very, very good at ruling out people for having amyloid pet positivity, which is really actually quite, quite striking. And in fact, if we adjust that optimized single cut point up just a little bit, you know, we could actually have a 98 to 99% negative predictive value on this test.

Racke, Michael K   27:18
So now can you tell us a little bit about dual cut points? Because I've heard a lot about that recently as people have started using panels with multiple biomarkers.

Stroh, Matthew A   27:29
Yeah, certainly. And you know, as I mentioned with these complex heterogeneous, you know, insidious diseases that are, they're slow moving and you know, they're putting people basically on a spectrum, you know, from being subjectively cognitively impaired to actually having mild cognitive impairment to, you know, developing full-blown Alzheimer's disease, dementia. You know this, this process can take quite a while.

So with these tests, it can be difficult to have single cut points. So the CEOI, which is the CEO initiative for curing Alzheimer's diseases ultimately has decided to issue guidelines that helped us to understand how we can develop dual cut point tests and have tests that are still high sensitivity, high specificity and give the clinician a good amount of you know, confidence in the results, but also minimize the individuals that are kind of in that unknown indeterminate intermediate area.

So you know, as you can see here with the probability of Emily PEP positivity, you know we have a high cut off and a low cut off that yields individuals that are positive, intermediate or indeterminate and of course negative with the CEOI saying that you know 90% of those tests that have, that have the intention of being confirmatory for Alzheimer's disease.

They should have a greater than or equal to 90% sensitivity and a greater than or equal to percent or 90% specificity, as well as have a less than 20% indeterminate rate.

So if a physician tests 10, you know, 10 patients in the clinic, they should have only two of those patients that actually come back as having an indeterminate result. The rest should have either a, you know, a high-probability or high risk or low probability, a low risk.

Racke, Michael K   29:32
So before we are using that same diagram that you showed earlier and basically with the same AUC, but now we've moved those cut those dual cut point. Before it was right in the middle because we said with a single cut point that was the best way to differentiate between the two groups, can you tell me a little bit more about, you know, you were talking about the 90% sensitivity and specificity?

Stroh, Matthew A   30:00
Yeah, yeah, certainly. So you know with the with the single optimized cut point you know on you know on a on a test that is less likely to be in you know in one of these really complex cohorts or disease populations.

Racke, Michael K   30:01
And how that's used here?

Stroh, Matthew A   30:17
You know, you may be able to end up with a test that has very high sensitivity and very high specificity, but as you know, these blood based biomarkers are being developed on, you know, diseases like Alzheimer's disease or you know, dementias in general.

You know, we've already talked about how complex and how there's overlapping, you know, etiology. So with, you know, with these dual cut point models, this allows us to basically control the sensitivity and specificity of the tests, still providing the clinician with, you know, the confidence.

In their positive or negative results while you know providing an intermediate zone or indeterminate zone for the rest of the patients that are just not, you know we're not quite confident in.

So this still allows us to, you know, rule in, rule out, but also still, you know, have that a bit of uncertainty.

Racke, Michael K   31:20
So then could you now apply this to some of our data on an intended use cohort in terms of how that you use those 90% sensitivity and specificity cut points?

Stroh, Matthew A   31:35
Yeah, certainly. Yeah. So for again with this, this is, you know, a cohort that we looked at again, you know, separated and stratified by pet status. And you know, over here on the left is the negative pet group and on the right is the positive PET group.

We have the plasma A beta 4240 plus P Tau 217 probability score. So what we what we figured out is that you know the more biomarkers and relevant biomarkers that we add the, you know the better we can increase the performance of these tests and we'll talk about that here in a little while.

But you can see here that the lower cut point is the cut point that provides us with greater than or equal to 90% sensitivity. And this one here at the top, the top cut point actually provides us with a cut off that provides you know, greater than or equal to 90% specificity and still yields an indeterminate zone that is optimized and minimized to, you know, to the best of our ability.

Racke, Michael K   32:43
So we've been looking at things where basically we had a single biomarker. How does it now work as you start adding multiple biomarkers and what does that do to our ability to improve accuracy of the test?

Stroh, Matthew A   33:00
Yeah. So you know, you know, as we showed early on in the presentation, overlapping, you know, biomarkers and complex etiologies and pathobiology of the diseases, you know, we're really looking to identify what a patient's disease profile kind of looks like, right?

What those, that pathology looks like in in the patient's brain by looking at their blood and you know, using Bayesian statistics and looking at the probability of being able to predict what's in their brain, we found that of course adding more relevant biomarkers actually helped us increase the performance of the test by reducing the actual indeterminate zone.

You know, so over here on the left here in A we see and with the probability of PET positivity with 4240 alone on the intended use cohort. So this is excluding healthy controls. We you know we see this wider indeterminate zone and if we look at just the probability of PET positivity with P Tau 217 here in B.
Of course we see a wider indeterminant zone again, but once we start combining biomarkers with the probability of PET positivity here in C with 4240 and 217 combined together in an algorithm.

You know, our indeterminate zone shrinks significantly down to about 1515% and with the probability of PET positivity with 4240P tau 217 and APOE 4 status, that's just telling us, you know, in a in a dichotomy.
You know, whether or not somebody is or is not an APOE 4 carrier, which is, you know, the greatest genetic risk factor for late onset Alzheimer's disease. This actually reduced our indeterminate zone, I believe down to about 10%, actually 12% for that one.

Now with the probability of PET positivity with 4240P Tau 217, and then of course APOE 4 allele count, which actually tells us the number of individuals, individual alleles that an individual is carrying, we're able to reduce that indeterminate zone down even further to about.
9 to 9 to 10%. So you can see a progressive decrease in the indeterminate zone and of course increase in the confidence in the test.

Racke, Michael K   35:34
So that data was looking at that cohort that you you described earlier of the patients that were very well characterized. How does it compare when you look at that to looking at real-world data?

Stroh, Matthew A   35:50
Yeah, great question. So you know we're a very large reference laboratory with a Quest Diagnostics. I think we have you know over 2000 PSCs, you know we have a high volume of orders that come in. So what we were really interested in doing is after we launched the panel looking at 4240 PTAL 217 as well as you know basically these individual biomarkers we were able to look and take about, I think it was over 4000 individual results from from patient results and look to see whether or not we were still getting similar predictions based on physician ordering patterns. So these are actually real-world results from physicians that are ordering the tests.

And we found that, you know, we actually had a better indeterminate likelihood of about 7% than we did in our actual development cohort, our intended use cohort of that we used to develop the test. With the high likelihood and low likelihood, you know, sitting very similar to what we originally observed. So this kind of shows that you know with physicians ordering the tests that we're able to kind of reproduce the numbers that we saw in our cohort that we used to develop the tests.

Racke, Michael K   37:16
So that's kind of interesting because that doesn't necessarily take into account what type of patient. Can you tell me, is there a difference like if I'm a neurologist that's ordering the test in clinic versus a primary care physician?

Stroh, Matthew A   37:36
Yeah, certainly. So you know this is something that the CEOI wanted to address in their in their guidelines. And you know I will kind of draw I guess your focus here over here to the confirmatory test in the top left.

You know this is where we were talking about you know the performance of the tests and the guidelines that they that they set out with a 90% sensitivity, a 90% specificity with a less than 20% indeterminate rate, right with the prevalence of amyloid positivity in the patient population kind of reflecting the clinical suspicion of AD.

You know, with a higher prevalence, obviously you're going to get a higher positive predictive value. You know, it's a lot easier to, you know, to shoot fish in a barrel if there are more fish in the barrel, right? It's a lot easier to hit the fish. So with a higher prevalence of about 80%.

You know this is something that we this prevalence that we'd see probably in like a memory clinic, memory disorder, specialty care setting and you know we get a high positive predictive value, but you know a lower negative predictive value and in turn with a lower prevalence at about 20% which is what you kind of suspect that you would see in the primary care setting. You can expect that the negative predictive value is going to be much higher at about 97%, with a positive predictive value being much lower.

Racke, Michael K   39:10
So if I'm a clinician and I have a patient in the office, how does that information that you just talked about affect the way that I would interpret the results that I get?

Stroh, Matthew A   39:25
Yeah. So, you know, the CEOI did a very good job of laying out, you know, different guidelines for different tests, right. So if there's confirmatory, there's triage, high specificity triage and triage tests. And they also did a really good job of outlining, you know, how these test results should be interpreted. Interpreted in the context of either a primary care setting or a specialty care setting. So you know in the context of you know with a with a with a biomarker positive result with a high clinical suspicion of AD, this is a 50 to 80% prevalence, most likely a neurologist office or memory care clinic.

There's, you know, this, this biomarker positive result means that there's a high likelihood of amyloid pathology. But if we look at those that actually have a low clinical suspicion of AD, this may be your primary care setting.

The amyloid pathology can neither be, you know, confirmed nor ruled out and this would help the physician decide whether or not to send on for further testing. So you know, at the primary care level, really this will help, you know, determine whether or not a patient is one that should be referred right away to a neurologist.

Or you know with a biomarker negative result. There are also other causes, secondary causes and potentially reversible causes of dementia that should be explored first on the on a negative result because again with a high a confirmatory test.

With a low clinical suspicion of AD, you know or amyloid PET positivity, those individuals, you know, have those tests have about a 97% negative predictive value, which means that if they're negative on the test, that's highly likely that they are negative on amyloid pet.

Racke, Michael K   41:19
So let me ask you this. If I'm a primary care physician and I'm considering a patient to refer to perhaps a memory clinic, is there an advantage for that patient if I order the test as opposed to waiting for the test to be done?
Down at the Memory Clinic.

Stroh, Matthew A   41:41
I mean there could be arguments for advantages, yes. You know the having the test results in hand, you know there there's some data recently that suggests that you know the availabilities blood based biomarkers is actually reduced the wait time to get into a neurologist's office or memory care clinic referral, but you know we'd also like to think of it from the primary care physician's perspective, referring to a neurologist or memory care clinic and you know what the exchange of information looks like and the time that is put into the first visit with, let's say, a specialist, right?

You know a specialist instead of you know having a patient come in and getting the patient history and then determining to do the blood test as the first step. You know with this with the PCP already providing the results from this test and having the results available to the neurologist then that just provides you know one step that that the neurologist doesn't have to take and potentially could get a patient either enrolled in a clinical trial faster or screen for a clinical trial faster or even on disease modifying therapy of which there have been two approved over the last few years.

Racke, Michael K   43:05
And isn't there data also to suggest that if you identify that person earlier using the blood based biomarkers that they derive, I'll say better therapeutic benefit as opposed to waiting till later?

Stroh, Matthew A   43:23
Yeah. So there have been a number of failures in the clinical trial space with anti-amyloid therapies, you know, due to due to different reasons. But you know, I believe that consensus is coming about that earlier is is better in these diseases, earlier intervention, you know earlier identification is better.

If you wait until somebody has, they're in your primary care physician's office and they're doing a mini cog and they score as being MCI or AD on a mini cog. That patient is probably, you know, well beyond what the DMTS might actually be able to help. So you know, early identification is best if we could find those patients that are showing changes in pathology, pathobiology very early. Even possibly, you know those patients that are SCI, subjective cognitive impairment or very, very early MCI, those patients are more likely to benefit from drug.

And I should also add not only you know drug, but there's more evidence that's actually been growing to suggest you know lifestyle intervention as well, you know diet, exercise something that we've all known about but the evidence really needed to pile up and this is this is growing.

Racke, Michael K   44:51
So I think that one of the things that I hope you've learned today is that we're really, I would say, redefining dementia, that we're not just using clinical characteristics of how the patient is cognitively impaired, but also using imaging and in fact using the biomarkers.

Stroh, Matthew A   45:18
Right. Yeah. Yep. So.

Racke, Michael K   45:19
Can you talk a little bit more again about how now in the future you might see this even improving more?

Stroh, Matthew A   45:28
Yeah, absolutely. Yeah. So this is where we kind of suspect that the field will ultimately go. You know, referring back to the very beginning, when we look at this, at this, this umbrella image that kind of puts these diseases into nice little buckets.

You know, we know that the reality is, is that, you know, a vast majority of patients are going to have some degree of, you know, mixed dementia or mixed pathobiology and etiology for the disease. So what we ultimately will would like to see happen and probably will end up seeing happen is you know the clinical approach using cognitive assessments, paper cognitive assessments in house. You know these are probably going to become more digital cognitive assessments.

And more in-depth AI driven and probably you know more comprehensive making the digital or the cognitive assessment earlier on in identification maybe at the PCP level even you know even more robust.

In addition to that, you know anatomically obviously we talked about looking at PET imaging and looking at amyloid pathology using PET imaging. But obviously you know individuals who have Alzheimer's disease show you know reduction in in brain matter, right, in both White, and Gray matter.

As their as their brain slowly, you know, dies and what we would like to be able to do and what we've actually demonstrated in one of our follow-up papers to our 4240 assay paper was that if we look at individuals that are on or that show as being positive on 4240 in plasma, but actually negative on PET, we can actually see using diffusion tensor imaging changes in the brain in the White, matter and Gray matter that suggests that there's already deterioration and loss of brain matter.

Beyond what we may be able to, you know, detect or see with, you know, amyloid PET deposition or even FGG PET with hypometabolism. So you know, we would really like to be able to combine of course MRI and advanced MRI.
Techniques including diffusion tensor imaging into the disease profile as well and all of this being accompanied with blood tests that evaluate the different biomarkers and ultimately potentially targets for drug.

And you know, what we would like to see this do is kind of redefine how we look at dementia and how we name dementia and how we, you know, I guess evaluate dementia for specific patients where ultimately at the end of the day, you know, we may end up having you know drugs that are disease modifying therapies for very specific pathobiology that can be mixed or used in combination such that we, you know, ultimately address the full pathobiology that's, you know, going on in the brain.

Racke, Michael K   48:56
So I think from today's webinar, I hope that you've gathered that Alzheimer's disease, like all dementias, really is multifactorial and that one really needs to consider not just the clinical aspects of the disease, but also the anatomical location of what parts of the central nervous system are affected, as well as the biomarkers that can help us define the disease.

Stroh, Matthew A   49:26
Yeah. And you know we also you know we we were able to discuss and you know how blood based biomarker testing is, it's scalable, it's an accessible tool that can infuse the timely insights into the into the clinical care pathway and you know as we develop more biomarkers and validate more biomarkers.

We are looking to combine them and develop algorithms to be able to differentiate different, you know, similar clinically presenting but distinct in pathology, pathobiology diseases by using multiple biomarkers in combination with each other.

And specifically for our cases, you know, for our purposes in the valuation of brain amyloid pathology and hopefully strengthen that predictive performance and reduce the number of indeterminants that we see.

Racke, Michael K   50:18
Yeah. So I think that brings an important point as we conclude the webinar and that is that, you know, the field has really dramatically changed with the first FDA approvals of anti-amyloid therapy, but I think as more treatments become available and more biomarkers become available.

You're going to see that the ways in diagnosing and caring for Alzheimer's patients is going to change rapidly and I think it's a very exciting time for patients with these disorders, because now we do have approaches that can at least have some effect on a disease that really was quite devastating and still is devastating, but the progress that's been made in the last few years certainly offers us with a lot of hope.

We thank you for attending and at this point we're happy to answer any questions that you might have.

Cuttic, Natalie M   51:26
Yeah. Thank you so much, Doctor Stroh and Dr. Racke, for sharing your expertise today. As Doctor Racke mentioned, we're going to move into the Q&A portion.

As a reminder, you can continue to submit your Q&A question or your questions into the Q&A feature at this time and we'll do our best to address as many as possible in the next few minutes.

So, so far there's really, there's just one question in the Q&A at the time at this time. Thank you for this very important information and this was early on in the webinar. She mentions you may talk about this later in the webinar, but wondering your thoughts on screening family members.

For APOE 4, my father and five of his six, five of his seven siblings had or have this disease and I want to know my risk as his 62-year old daughter. So I know you mentioned the advantages of adding on APOE 4 and the allele count. So if you could just sort of reiterate the advantages there of screening family members for this disease with a clinical history in their family.

Racke, Michael K   52:33
So I can tell you that, you know, this gets into an area that's perhaps a little bit controversial. The American Academy of Neurology for sure has basically suggested that if you're asymptomatic, you should not be evaluated with these blood based biomarkers because there is no approved therapy at this time for somebody that has positive biomarkers but is normal. So right now all the appropriate use guidelines are for people who have cognitive impairment.

One of the things that I found interesting, I thought that maybe at a memory clinic that people were automatically getting screened for Apo E4 and they're not. And typically it's, you know, there is an increased risk. If you have one allele of Apo E4, your risk goes up about four times. If you're positive for both alleles with APOE 4, that goes up to 15 times. And there have actually been some recent publication to suggest that you're if you live old enough, you're going to get Alzheimer's disease if you're homozygote for APOE 4.

The interesting thing is, and why I say it's a little bit controversial, there are studies that are going to read out next year where patients who are at risk for the development of Alzheimer's disease typically in their 50s and they're going to see whether the treatments at that initiated at that time.

Actually, by removing amyloid can slow down the occurrence of symptoms for Alzheimer's disease. And if that, if those studies read out positively, I think a lot of what we, I say here is going to one of the things that Matt brought up though that is very important to recognize and that is that uh Wellness interventions that recently the pointer study but as well the finger study and a number of other studies suggest that if you live a healthier life, lifestyle, you can actually reverse some of these biomarkers.

And when I say to people, you know to physicians, that while the American Academy of Neurology says you shouldn't, I'm sure some of you saw the CNN special with Sanjay Gupta, who's a practicing neurosurgeon, so I hope he's cognitively normal that he had blood based biomarkers assessed while he was on this show.

And that is because there are people in the field who think it's important to know what your biomarkers are and that if they're positive, that gives us a window of opportunity to do things like Wellness interventions.
To try to reverse some of those things that can contribute to further amyloid deposition and then finally the formation of neurofibrillary tangles.

Cuttic, Natalie M   55:33
Thank you very much for that. That's very helpful. We have another question in the queue. What is on the horizon for testing autoimmunity in dementia?

Racke, Michael K   55:50
Matt, do you want to take or do you want me to take that?

Stroh, Matthew A   55:52
No, that's a that's a great question. It's also a very difficult question. You know when we say autoimmunity in dementia, I I'm, wondering if we're talking what we're what we're specifically referring to.

You know, I don't know if we're talking specifically about maybe autoimmune reactions to disease modifying therapies or whether or not there is, you know, autoimmunity that drives dementia. So this is this is kind of a difficult question to answer. Um, Mike, do you do you have any uh?

Racke, Michael K   56:31
Yeah. No, I mean, I think there's how do you, I'm not exactly sure where this question was going, but I'll use as an example. You know, there's a lot of interest right now in vaccine development to try to prevent Alzheimer's disease and some of you probably in the audience may have heard that there were vaccines to amyloid with the idea that you would make antibodies to amyloid and then you would constantly be removing them and that would be of benefit. But the interesting thing is that what happened.

Stroh, Matthew A   56:52
Mhm.

Racke, Michael K   57:05
Is that in addition to making an antibody response against amyloid, the patients also made a cellular immune response against amyloid. And so these patients developed autoimmune encephalitis and I think some of that is what we see.

We didn't really talk today about something called Aria or amyloid related imaging abnormalities and that has to do that when you give the amyloid therapy, you're having antibodies react with amyloid and that is triggering the immune system to help in terms of clearing and when you think about autoimmunity in the nervous system, that gets to be really interesting.

There was recently a paper published by the Wash U group that showed that patients with multiple sclerosis were less likely to develop Alzheimer's disease, and they think part of the reason for that is because you've got this immune activity going on in the brain that's helping to clear away amyloid, which and turns out to be beneficial even though that immune activity is detrimental in. In terms of causing the myelinating events that lead to the attacks in multiple sclerosis. So I mean I think that this is an area where there's a lot of interest and the others, I don't know if I say the other aspect of that is that.

You know, we talk about some of these P Tau's as being overrepresented in Alzheimer's disease. But for example, there was a paper from Oscar Hanson's group demonstrating that patients with progressive multiple sclerosis who were GFA positive also were positive with much lower levels than an Alzheimer's disease of P Tau 217. And so, I think the other thing that that that tells you is that a lot of these biomarkers, they're not necessarily specific for a disease.

They're specific for a neurodegenerative process. And I think that, you know, that's an area that's going to also have a lot of expansion. I can tell you we were just at the recent Alzheimer's Association International Conference.

And you're now seeing biomarkers for a number of other neurodegenerative disorders that typically you don't see, for example, in Alzheimer's disease, but are present in things like cortical, basal and cortical nigral degeneration. So you know, as we mentioned earlier, the field is really changing rapidly and I think that's going to lead to a lot of new, exciting developments.

Cuttic, Natalie M   59:48
Thank you so much. Unfortunately, we are at time and there were a few other questions in the queue that we would love to follow up with you on those. And when we send the recording out, we'll put a link in there to request aconsultation.

If you'd like to discuss any of these topics further, we'd be happy to connect you with one of our experts. So thank you to everyone who submitted questions and a special thanks again to our speakers for such an engaging discussion.

Stroh, Matthew A   1:00:17
Absolutely.

Cuttic, Natalie M   1:00:24
And you know, on behalf of Quest Diagnostics, thanks for joining us. We hope today's session gave you a valuable insight into blood biomarkers and their evolving role in AD diagnostics. Have a wonderful day and we look forward to connecting with you at our next webinar.
Thank you.

Stroh, Matthew A   1:00:42

Thank you.

stopped transcription