INmune Bio Inc. (INMB) Earnings Call Transcript & Summary

Good afternoon, and welcome to the INmune Bio Virtual KOL event. [Operator Instructions] As a reminder, this call is being recorded, and a replay will be made available on the INmune Bio website following the conclusion of the event. I'd now like to turn the call over to your host, R.J. Tesi, Chief Executive Officer of INmune Bio. Please go ahead, R.J.

Thank you, Tara, and thank you, everyone, for joining us. As you know, I'm R.J. Tesi, the CEO of INmune Bio. And with really great pleasure, I introduce this webinar on why EMACC or Early Mild Alzheimer's Cognitive Composite is the optimal tool for measuring cognitive change in early Alzheimer's disease. We are very fortunate to have Judy Jaeger, who's the Founder and Principal of CognitionMetrics, a consultancy that helps pharmaceutical programs design and deliver clinical trials for drugs that impact cognition and behavior. Judy Jaeger led the team that developed EMACC, and that's the tool you'll be learning about today. Judy maintains her academic ties with academia as a professor at Albert Einstein College of Medicine. Also speaking will be Sarah Barnum. She is a Director of Clinical Sciences at CognitionMetrics. Sarah is a Board-certified clinical psychologist with deep expertise in neuropsychological testing and holds an academic appointment at Oakland University William Beaumont School of Medicine in Michigan. C.J. will also be speaking, C.J. Barnum. You know C.J., he's the VP of our CNS Drug Development program. And yes, you noticed that C.J. and Sarah have the same last names, they are married. With that, I will turn it to you, C.J.

Thanks, R.J. And I'm really, really excited to have Judy and Sarah here to talk about something that is probably the biggest question that we get surrounding our study. And I'm hopeful at the end of this, you will be as impressed by the diligence that they've done to develop an endpoint and implement an endpoint in a trial that I think is really critically important. So I'm going to briefly go over the objectives here, and then I'm going to pass it off to Sarah. So what we're going to start with is what is early AD and how it's currently assessed. We'll talk about the conventional assessments, why they're either inadequate or suboptimal for early AD trials. And then we'll talk about how to overcome this. And of course, this is where EMACC comes in. Many of you have asked about the regulatory pathway for EMACC. So Sarah will talk about that in greater detail. And then Judy is going to show you how EMACC and the data that we've been able to generate and has been generated outside of INmune Bio really meets the requirements for approval. So with that, I will pass it off to Sarah.

I'd like to start by talking about the Alzheimer's disease continuum. This is really a disease that varies widely across the course of illness, starting at the left-hand side from our preclinical patients who aren't yet showing clear indications of concerns all the way to the right-hand side where we have our severe AD patients who have significant impairments related to their Alzheimer's disease. In AD02, an immune study, we're really focusing on early Alzheimer's disease. And when we talk about early Alzheimer's disease, we're talking about patients at the stage of illness that are qualified for MCI, mild cognitive impairment, or at the early end of mild dementia. So these are patients who are recognizing change in their cognition. They're noticing memory problems and they're starting to impact them on a day-to-day basis for some of their skills, but they're still independent for most of their basic day-to-day living. Research in Alzheimer's disease has really been primarily within that mild-to-moderate stage of illness. And the main accepted cognitive endpoints in the trials that have been run so far, there's 2. One is the ADAS-Cog, that's the Alzheimer's Disease Assessment Scale. And one is the CDR, the Clinical Dementia Rating scale. The ADAS-Cog is a combination of performance-based measures. We're asking patients to complete a task to see how they can do on a specific task. And then some clinical ratings. The person who's administering the test is making some decisions about whether or not there are impairments in certain areas. The ADAS-Cog was developed for that mild-to-moderate stage of disease. CDR is a bit different. It's a clinical rating scale. So it is, again, trying to assess the level of symptoms that a patient is showing. It was really created to determine does a patient have Alzheimer's disease and at what stage of disease are they based on the symptoms that are being exhibited. Now these endpoints are generally used in cognitive testing in AD trials, mild-to-moderate disease. But as the research has gotten better, we've started to intervene earlier and earlier into the disease. So because these are the approved endpoints and the measures that are available, they've been used in this early AD stage as well. Next slide, please. So again, we're showing you that Alzheimer's disease continuum. I mentioned CDR as a staging tool, and you can see the global ratings there that correspond to preclinical disease, early Alzheimer's disease and then later in the disease course. And then additional measurements have been developed to assess cognition at different points in the stage of illness. And so we've got measurements for preclinical. We've got measurements for mild-to-moderate disease. We've got measurements for severe disease, but you'll notice that there's a gap. And that gap in measurement really falls within the population that we're interested in looking at in that early Alzheimer's disease population. You may recognize, based on why these tools were developed and the type of tools that they are, that there are some drawbacks to using the ADAS-Cog and the CDR in this early population. The ADAS-Cog, again, was developed for more severe disease. And this creates a problem. When we have a measure that doesn't match the stage of disease, we can run into a couple of problems. And one of the problems that we run into is called floor effects or ceiling effects. So a ceiling effect is when you give a test and everybody performs perfectly. So if you perform perfectly on an item or on a test, we can't measure any improvement in your skills. You're at the top, there's no way to measure change up. A floor effect is if you don't get any of the items correct or if you're not able to do the task at all, we can't tell if you get worse over time. So that floor and ceiling effect really impacts our ability to measure change in cognition over time. And that is a big problem with the ADAS-Cog in our early AD population. 70% -- actually more than 70% of the items on this measure, so 11 of those 14 items, don't work in early AD because of floor or ceiling effects. And so that really only leaves 3 questions out of the entire tool that can measure early AD symptoms. So this measure does not allow us to detect change in early Alzheimer's disease. It really isn't a match for this stage of illness. Next slide, please. Now I want to focus on the CDR. As I mentioned, the CDR is a diagnostic rating scale. So it was developed to determine diagnosis and stage of illness. There are 3 main sources of information that we gather using this CDR. The first person is the informant. So this is somebody who is -- someone who's in the life of the patient. It may be a family member, a spouse, an adult child, a neighbor, a close friend that we ask for information about the patient's symptoms, about their abilities and about change over time. We also ask them to give us some information about events in that person's life so that we can then ask the patient about those same events. So one way of getting information is by asking an informant. A second way that we get information about a patient's skills are we ask the patient. So we may ask them, do you experience difficulties with your memory? And then we also ask them to tell us about the events that the informant has shared. And we have them complete some specific activities so that we can look at what -- whether or not they can do certain tasks and what that tells us about their abilities. Then the clinician uses the information that they've gathered from the informant and from the patient to make a decision about what is the right score for this patient. And they'll assign a staging score. So you'll see in this little box with the green highlights here that they're going to give a score from 0 to 3 for are there no symptoms present, are there some questionable changes, are there mild changes, moderate concerns or severe dementia? And those -- that is what gives you that global score that you saw along our arrow a minute ago. Now this presents some significant challenges. Using a diagnostic rating scale in this way, there's a lot of noise, there's a lot of variability that's introduced into that final score. The informant creates some noise. We're asking somebody based on their impressions, their observations, have they seen these things? What is their opinion of how the patient is able to do certain skills? We might ask them, can they remember a list of shopping items? And they usually remember. Can they remember sometimes? Are they not able to remember? And we're really relying on them to have good information to have insight into that information to communicate it well to us. The patient also introduces some variability. We're expecting patients that we know have decline in memory to be able to understand and have insight into and be able to communicate their level of impairment as well. And you can see that there are some challenges in both of those areas. And then finally, the clinician has to synthesize this information. It is not quantitative. This is a judgment based on the information that they have. So that CDR-global is made in this way. You'll hear also about the CDR Sum of Boxes. And this is how they've repurposed this diagnostic rating scale for use in clinical trials. And basically, what they do is they take the score, that 0 to 3 score, across 6 different domains, domains like memory, judgment and reasoning, daily living skills, and they add them up. So that's your sum of boxes, as they're adding the CDR score from each of the domains to get a total score that's going to range from 0 to 18. So again, a diagnostic rating scale with quite a bit of noise, potentially some variability, and then we're getting a total score based on that information. Next slide, please. I want to focus a little more on the limitations based on the informant. Recently, in fact just this year, some new research came out looking at whether or not the type of informant, who you're asking the questions about the patient, does that make a difference in terms of the ultimate CDR score that's given? And what they found is that there are big differences, and there are a variety of sources of influence here. So the gender of the informant, so a female versus a male informant, if it's a female informant, you're going to end up with a higher CDR score. The relationship between the informant and the patient really matters as well. An adult child of a patient is going to report more symptoms and result in a higher score than, say, the spouse or other caregivers. The frequency of contact is very important. The more time that the informant spends with the patient, the higher the score on the CDR. That's important as we're thinking about planning clinical trials and who are we using as the informant. And then ethnicity also matters. And as we think about diverse clinical trials, if you've got a child informant with specific ethnicities, you're going to end up with higher or lower scores. So a lot of variability influencing the ratings regardless of the disease severity. And that creates problems for really being able to pick up a signal in clinical trials. Next slide, please. I want to give you a sense of the magnitude of this difference, like how significant is it who the informant is. And we've got some good data on that now. Looking at that arrow across here, this is really the magnitude of the effect of different factors on the ultimate CDR Sum of Boxes. You'll notice that the relationship and the frequency of contact, those red circles there, really rival the level of change that we're capturing in treatment effects. So CDR is almost changed as much by having a different relationship of the person informant as we get in a lecanemab trial. Similar, the frequency of contact of a patient rivals the treatment effect in both lecanemab and donanemab trials. And so this can really add a lot of noise, add a lot of concern in our confidence about using CDR as a measure of change because the informant can really affect the data as much as a potential treatment can affect the data. Next slide, please. So based on all of these limitations of these measures, the fact that they were not developed to measure cognitive change, that they're not a good match for the level of disease that we're studying, we really don't believe that they're an appropriate endpoint for early AD studies. They're not what we call fit-for-purpose. And so we have a real risk related to using these measures for clinical outcome measures in early AD trials. And that risk is that we miss the effect of a drug on cognition in a clinical trial, and that is a very concerning risk. It's a problem. It's a problem to solve. And there really are 2 ways to potentially solve that problem. One, is to have longer, bigger studies. So if we have a study that includes really large number of patients and we follow them for long periods of time, we can watch them move from one stage of illness to the next, and we can get a sense of some of that change, but it really requires very extremely long, large bulky trials. That's what other companies have used so far to solve this problem. We think that there's a better way. We believe that developing a more -- a better measure, a more sensitive measure of cognitive change that's a fit for this early Alzheimer's disease population is the way to solve this problem. And I'm really excited that Judy is here to present the EMACC to be able to describe this tool to you and talk to you about why it solves some of the problems that we've seen in the other tools used in early AD trials. So I'm going to hand off here to Judy and let her continue with you. And Judy, you are on mute.

Thanks, Sarah. I was approached, gosh, close to 10 years ago by several pharmaceutical companies to help them with the problem of what should be the optimal cognitive endpoint in early AD trials. And I approached it with data. I believe data is the best driver for decision-making in clinical drug development. We developed then the EMACC. What is the EMACC? The EMACC is a combination of objective validated tests that are used routinely to measure cognitive changes that occur in early AD. They are used in memory clinics all over the world. Many of these tests have been around for 50 to 100 years or the test paradigms. They're very common. They're very familiar to neuropsychologists and clinicians all over the world. So that's what the EMACC is. And I'm going to talk to you about what problem it solves. Sarah already related to you the sort of spectrum and where measures are suitable. And she demonstrated to you the gap, and the problem that floor and ceiling effects cause in the use of the ADAS-Cog for measuring cognitive change that it becomes very insensitive. You're relying on just a handful of items, and hence, you need really big trials and longer follow-up to be able to detect a drug effect. So that is the problem we sought to solve, and that is the gap that the EMACC fills. In the next slide, I'm just going to summarize for you what we did. So I gathered up colleagues from -- who represented scientists, who represented and were investigators in several longitudinal aging studies. They're listed under the methods box there. So people who worked on ADNI, AIBL, people from Washington University and the Mayo Clinic Study of Aging, all had longitudinal data year-on-year repeated cognitive testing in individuals who were very well characterized, including with biomarkers. So confirmed amyloid positivity in that early stage that's actually useful. And we empirically asked the question from the large neuropsychological test batteries that were administered on a regular basis over time in the population of interest here, those who had mild cognitive impairment or early AD and had the biomarker, which combination of cognitive tests shows the steepest cognitive decline across the 4 cohorts? We only considered tests that were validated, were well known and widely used by clinicians and would be suitable for use across cultures because the objective was to develop a composite battery that would be used in global clinical trials in early AD. So it's that combination of tests that constitutes the EMACC, and the next slide will show you what they are. And I want to give you a feel for what this is. This is quantitative. It's not a rating scale like the CDR and many of the items on the ADAS-Cog. It's not a person's opinion. I call this a stress test for the brain because similar to stress test for cardiac function, we're asking the organ, the brain, the person to work at something, and we're measuring how efficiently that work is accomplished. And so the values are things like how many words you can remember if I ask you to memorize a word list. How many digits can you repeat when I do -- when I read you digits kind of like a phone number. How many words can you produce if I give you a task like tell me all the animals you can think of. And then we have some paper and pencil performance tests that are illustrated below. And we measure how many are correct or how many seconds it takes to complete the task. And then we combine those using statistical methods to come up with an average of each of those scores, and that constitutes the EMACC. So it's completely objective. There are no opinions involved. Next slide, please. Does it work? Well, sure does. We actually averaged the change -- the measure, the magnitude of change in that 4 sample, those 4 individual aging cohorts. And we're able to show that even after only 2 years, you could robustly show decline even in the earliest in the MCI stage of this disease. So in summary, EMACC is the first clinical measure for early AD. It was custom designed for early AD. It's objective and quantitative, no rater bias. It's specific to AD-related decline in cognition because we did have a control group. We controlled for non-AD-related change such as learning effects because we had a control group who did not have AD. All of the tests are validated, commonly used and understood by neuropsychologists all over the world. By definition, they didn't have floor or ceiling effects. We excluded any measure that did. And this measure has actually been adopted in a number of trials. We are not the first with INmune Bio. It's been used to my knowledge or is in current use in 5 trials in early AD by 3 different pharmaceutical companies. Next slide. So I'm going to pass it over to Sarah because the question I now get all the time is, what do the regulators think about it? Well, the regulators haven't had the opportunity to opine because typically that will happen once there is a signal with a drug. So there is a strategy, and I'm going to pass it on to Sarah to discuss the regulatory strategy for the EMACC.

So I want to start by mentioning that the FDA is really open to the development of new measures for clinical trials, for outcome measures. And they recognize the need. They put out specific guidances to help companies, to help researchers to develop what we call fit-for-purpose clinical outcome assessments. And this is pulled directly from their website. There are a number of different guidances that really work as a road map or a checklist for how to develop a measure that's fit for purpose and that's appropriate for certain patients in certain disease stages. And so we've followed their guidance. Don't let this slide overwhelm you. I know that this is a very busy slide. I just want to give you a sense of the level of detail here and to draw your attention to a couple of specific terms that are put forward by the FDA and that are check boxes, questions we can answer with the EMACC. So if you look at the yellow highlighted areas, you'll notice some very important terms when you're developing a fit-for-purpose patient-focused clinical outcome assessment. One is the concept of interest. So what are the most important areas for patients? And are you addressing those? The second is what type of clinical outcome assessment are you using that's going to adequately capture those important concepts? Number 3 is the context of use. Are you using it in the right patients for the right reasons with the right -- and with the same specific plan? And is your clinical outcome assessment fit for purpose? Do we have information to show that it is reliable and it is valid in this population that you're evaluating? So if you go to the next slide, those highlighted terms are now at the left-hand side of this box. And I want to talk with you about how EMACC maps on to the specific areas that the FDA is asking us to evaluate when developing a new clinical outcome assessment. So you see there on the left concept of interest, clinical assessment type, context of use and fit for purpose. We really map on to their specific framework, and we can check each of these boxes with the EMACC. So thinking about concept of interest, it's very important to the FDA that we are using patient-focused outcome assessments that we're addressing the symptoms that are most concerning to patients, caregivers, clinicians, advocacy groups, and the EMACC does that. Cognitive decline is the greatest concern for patients and families, and the skills that we're evaluating using the EMACC, memory, language, executive functioning types of skills are the exact types of skills that they're recognizing and reporting as their areas of specific concern. So we're matching that concept of interest well with the EMACC. Clinical assessment type is also very important because we need to have the right way of evaluating those concepts of interest. For -- when we're specifically thinking about cognition, about skills, a performance outcome measure is really the best way to evaluate skills in a patient, to ask the patient to do specific tasks to show those specific skills. It's advantageous because it's not relying on input from a rater. It's not relying on input from an informant. It's a really -- again, that stress test for the brain. We are asking the patients to show us the skills. And we have very standardized ways of administering those measures and very standardized ways of scoring. So no matter whether I'm giving the test or Judy is giving the test or someone else, it's going to be administered the same way and it's going to be scored the same way. There's not bias introduced that's quantitative. So that performance outcome is very important. Context of use is making sure that our performance outcome measure for cognitive decline matches the stage of disease that we're using it in. So the EMACC was developed specifically to fill that gap to measure change in cognition in early Alzheimer's disease. And that change is really important. We're not just staging. We're looking at change over time. We've got the measure that data shows is able to recognize change over time. And then ultimately, a measure has to be fit for purpose. So it has to be reliable and valid in the population that you're using it in. As Judy mentioned, the EMACC is made up of components that are widely used and well validated over long periods of time globally. And so those component measures have already been validated. And there have been additional steps taken to make sure that as a whole, the EMACC also is fit for purpose. And Judy is going to actually show you some more of that data in just a minute because we've got really nice information to show that we've got good validity, good reliability in those early validation studies and in AD02 in the trial that's ongoing. And so EMACC maps on to the FDA's clinical outcome assessment framework. We can check those boxes. We can use their terminology. And because of that, we believe that there's a high likelihood that the FDA will agree that this is the right tool to measure cognition and change in cognition in early Alzheimer's disease. This is just a graphic kind of representation of the same concept I've already mentioned to you, but we're mapping it onto a specific template that the FDA has asked for clinical outcome assessments. And so you'll notice at the left, again, we've got those specific health experiences, those concept of interest that are important to patients. And to the right, we have the EMACC, the neuropsychological measures of memory, language, executive functioning. And they meet in the middle at that most important concept of interest, the decline in cognition. And that conceptual model matches our measurement model. It's fit for purpose in early Alzheimer's disease. And so this is one of the things that we'll take to the FDA to prove to them that this is really the right tool in this patient population. So I'm going to pass back now to Judy and let her share some of our recent -- the recent data that was just presented so she can show you how the EMACC is fit for purpose.

Yes. So we were really delighted that INmune Bio was prepared to allow us to take a look at AD02 data. This is only baseline data, of course. We haven't looked at any outcome data yet. But you can get a good feel for how well a tool is working by having a look at the distributional properties and other things. And one of the things you'll notice in the first box there is that the EMACC is performing absolutely as designed with respect to distributional properties. There are no flooring-ceiling effects. You can pick up worsening and you can pick up improvement in any patient in this study. And that's going to make it sensitive to change because you can pick it up. You're not hitting a floor or ceiling. We were really over the moon about the second graph because what we've done is, in the AD02 trial, and I typically advise clients, pharma companies to do this, and it's becoming more common, to administer cognitive tests during the screening period as kind of a practice, a familiarization session because you don't want patients coming in on the first day of a study just before dosing, feeling anxious and uncertain about what's going to happen. It makes us worry that we might not be getting a really accurate assessment of their cognitive abilities. So we typically give the entire battery as if it were for the trial as a sort of practice during the screening period. And then, of course, we give it again at baseline, and that forms the basis for the change score when we look at drug effects. There's about a month interval between these 2 measurements. And as you can see, they're correlated at a level of 0.93. For any behavioral measure, this is probably the best I've seen. I mean, there are laboratory blood tests that have correlations in this range or not even as good as this. And what this is telling us is we really know that it's reliable and the lack of rater bias is fully also captured and evident in these data. And finally, there's a kind of validation, which we call criterion validation, which basically says, are 2 groups of patients who are even subtly different from one another, distinguishable on your measure? Well, these are the AD02 patients. And on the left -- that left dot, those are patients who were classified with respect to stage as 0.5 at the time they were recruited. So they're randomized into the study at the stage of 0.5. And on the right, you see those patients who were randomized with a CDR stage, CDR-Global rating of 1. There's no overlap. This is a tool that is robustly distinguishing patients who are subtly different from one another. Believe me, the decision as to whether someone is a 0.5 or 1 just when they're in that cusp is actually a really difficult one. It's not a terribly precise thing. And yet these subtle differences are robustly detected using the EMACC. If it means anything to you, the effect size of the difference is 0.87, a very, very high number. So that further confirms the sensitivity and validity. And we're going to be taking these data to the FDA as part of our package. And these data make me even more confident when we add it to the original validation data from those 4 aging cohorts, even more confident that the FDA will view this favorably. And on the last slide, I just want to conclude. So it is our opinion that the EMACC is the right tool for measuring change in early AD. It's quantitative. We're measuring behavior in the same way we measure any other physiologic event. It is a stress test for the brain. We can demonstrate that it's sensitive to the effects of disease progression and hence, it's going to be sensitive to the effect of a medication that impacts the disease. And what is the great advantage? Well, you could use the other less than fit for purpose, less sensitive tools that Sarah described. But if you use a more precise tool, you can run a trial with fewer patients, shorter duration and substantially lower cost. And what's most important to me is that means we're going to get -- bring drugs to the market faster and just as important, abandon those drugs that don't actually have benefit quickly so we can move on. It reduces the patient's burden and incurs significant overall cost savings.

Thank you, Judy and Sarah. I just want to take a moment to thank both of them. I mean, excellent presentation. And I hope you guys can all see why we are so enamored with EMACC. And I think to sort of reinforce what Judy said on this slide, when you're a small biotech company and you don't have the luxury of doing 18-month studies with 5,000 patients, I'm sure you can imagine the cost of that. And so some of the failures that we've seen with biotech doing smaller studies and fewer -- with less time, we believe, are attributed and can be attributed to the endpoints that they are using. And this is why we have such incredible confidence in what we're going to see with AD02. So with that, I will turn it over to R.J. I know there are some questions. And of course, we'll be happy to take those. Thank you.

Yes. Thank you. Fantastic delivery of some very interesting information. I'm going to ask a few questions that have come in on the internet, and then we're going to open it up to questions directly from you. So Judy, I mean, you've made -- you and Sarah have made a very compelling argument that this is a fit-for-purpose stress test of the brain for patients with early AD. And why haven't we seen more of it? Why aren't more companies using it?

Well, I mean, there are a number of reasons. The field was struggling with this middle ground for quite some time. And one of the advantages of using a tool you know, if you're a big pharmaceutical company, is you can actually estimate you have data. And so CDR, there's tons of data on CDR and ADAS-Cog. And so they're going to go with what they're familiar with. This is new and in fact, wasn't even available when the earliest trials began. This wasn't around when the aducanumab trial kicked off. So I think that's the major reason. And then it's a conservative industry, people follow one another.

Thank you. A question for Sarah Barnum. I was very intrigued by the figure that showed that the variance of some of the elements of CDR Sum of Boxes and was encroaching on the same kind of effects you see with the drugs, the anti-amyloid drugs that have been approved. If you were to guess that people were to use EMACC as a tool to show clinical benefits using those trials, would the results be about the same as what we see? Do you think they might show more benefit?

We believe that the EMACC is positioned well to pick up change -- have a more sensitive way of picking up change. And so we think that we're going to have a much stronger effect size and an effect size that we can feel more confident in because there are fewer factors that are impacting the change in cognition on our measure. So the change we see in EMACC is going to be related to the intervention. We've shown that with our test -- retest that graph that Judy showed, the correlation between screening and baseline. If you test somebody twice without intervention, their score is virtually the same over time. So if we see change, we can be confident that that's related to the intervention or to the progression of disease, not some of these other factors like who you're asking the questions to or what their opinions might be.

I'd just like to add, or if the inform changes in the course of the trial that does happen, where it was the spouse and then the spouse becomes unavailable and the daughter comes in instead. And now you're really comparing apples and oranges. And with respect to effect size, it's a technical term, but the amount of noise around a signal affects the effect size. If you can reduce the amount of noise, even with the same distance change, it doubles the effect size as a statistical matter.

Yes. Thank you. Judy, I mean, recently, there was a company or a slide that was presented or shall we say a study that was presented where the trial didn't work, and there was a comment that the placebo group didn't behave as expected. Does EMACC control for placebo effects any better than more traditional methods?

Well, yes, it's one of the great advantages. So what's a placebo effect? A placebo effect is a combination of many things, but it's largely expectancy bias, right? I have something wrong with me. I join a trial. I hope I'm on a medicine that's going to make me better. I'm sort of expecting it because why would I be doing all of -- making all of these efforts and taking this drug if I didn't think it was going to make me better. That expectancy bias doesn't only impact the patient when you're using the CDR. It impacts the informant who's also hoping the patient is going to get better. And frankly, it impacts the clinician who's hoping the patient is going to get better. So these expectancy biases are much more likely to impact CDR. I mean efforts are made to try to reduce it, of course, but it's unavoidable to a degree. When you measure performance in the way you do with cognitive testing, there's much less impact. It's not 0, but it's much less impact. So I think that we're much less likely to see placebo responding on EMACC. And certainly, we didn't see anything like that in these longitudinal aging studies. I mean they get worse. So we'll see, but I'm fairly confident EMACC will be immune or close to immune to the…

One more question until I -- and then I'll turn it over to Tara for direct questions from the participants. This one is for C.J. C.J, INmune Bio has adopted this precise purpose-built cognitive scale that allows you to do smaller, shorter clinical trials. And as you know, we've been -- we on Wall Street have been quite worried about that. But are there other elements of the clinical trial design that also play to your favor?

Yes. I mean I think the simple answer to that is the enrichment strategy. I mean I think this is a biotech strategy that has been shown time and again to be successful. I mean cancer is the best example of that. Success rate in clinical trials just increased significantly once they started enriching. It's not something we've done in neuro diseases. Part of that is because traditionally speaking, you can't biopsy the brain, right? So we are fortunate enough by targeting inflammation. We have the tools to measure inflammation and by enriching for the patients that are most likely to respond to the drug because it matches the target of a drug and the pathology of the patient. That makes a difference. The other thing that I want to add is when looking at EMACC, when we talk about fit-for-purpose, what we typically see, the way we describe Alzheimer's disease is that inflammation is a phenomenon that actually comes up a little bit early. And the thought is that in the earlier stages of the disease, inflammation is driving the development and progression of disease. And if that's the case, then what you would expect is a measure that is capturing change in that early stage should be more sensitive to capture change. And this is exactly what we see when we powered the study is that the patients that have biomarkers of inflammation on the EMACC, the EMACC was almost twice as good as capturing change in those patients than those that didn't have inflammation. So not only do we have a fit-for-purpose, but it matches the MOA. And really, when we talk to people and they see my excitement, this is why I'm excited about this.

Well, thank you. Tara, questions from the phone lines or whatever you call it in this day and age.

So our first question comes from Tejas Wein at Raymond James.

This is really informative. I guess just my first question would be, you've kind of alluded to before, clinicians on the trial are having a really good time using EMACC and have positive feedback on it. I guess if you could just go into any further detail, just any qualitative things you've heard on how they're doing with it and feedback? And then I have a few follow-ups.

Sure. So it's objective. We train the sites so that everyone is giving it in exactly the same way, but it's also commonly used. So every one of these measures is familiar to the neurologists who are taking care of these patients. So that's a great plus. They're comfortable.

And that's different than some of the historical measures. So in clinical practice, the ADAS-Cog is not used. This is something that people who work in memory clinics, who evaluate and assess and monitor change in patients, don't use that tool. They use the components of the EMACC and other neuropsychological tests clinically. And so they're very relevant and understood well.

Yes. And then another question that I had. You mentioned kind of testing patients first at screening and then when they get on the trial to kind of give them a little bit of practice and familiarity. How are you striking that balance between giving patients some familiarity with the cognitive metrics and also preventing them from becoming too familiar and kind of knowing what the questions are going to be? How much variability in those individual items is there?

Right. So you're talking about learning effects that people worry about. We don't worry very much about learning effects, the interval. I mean, learning effects are things you worry about when it's a few hours apart. But over weeks, we don't typically worry about these. Also, this is Alzheimer's disease. If there's a little bit of a learning effect from prior exposure, if anything, it will increase our signal in a clinical trial. We can demonstrate that these patients decline even when you're giving these tests at the frequency that we're giving them. Having said that, we do have some alternate forms so that, for example, the words on the word list are different every single time. And there are different versions of many of the tests that we give.

Yes. And then I guess just my final question is, do you see yourselves as kind of as INmune Bio bringing this EMACC measure to kind of the forefront of Alzheimer's? And how do you expect to see the community kind of adopt this measure more? If you're able to show a good effect in your trial and really validate this as a metric, where do you see that going?

C.J., I'll let you…

Right. Well -- okay. I was just going…

No, go ahead, Judy.

You guys aren't first.

You go ahead, Judy.

I'd like, Judy…

AD02 is the fourth trial, so -- but go ahead, C.J.

No, I think Judy is right. And I think Judy actually has more to say on this than I do. But from my perspective, I think it's -- what's clear when I go to these meetings and you talk to people, it's very clear that we don't have the right tool. And it's almost like the world is waiting for somebody to do something about it. And we're doing something about it. And so I think what's going to happen is once we demonstrate this, we bring this to the FDA, the FDA gets their blessing. I think it's going to completely change the way it's used. I think the uptake is going to be very high because remember, the biggest issue that I see now, and I see this in -- at these meetings is, what folks will do is, they'll do a 6-month study and they'll use the CDR, the ADAS-Cog and then they go, well, it didn't work. And they say 2 things. One, we didn't go out long enough. And we didn't go out long enough because the placebo control group didn't decline. It's because they have the wrong measure. And so I think there's going to be a lot of light bulbs that go off once this comes to the forefront. And I think it's going to make a huge difference for the field and it's going to make a huge difference for investors because of what it's going to do for bringing drugs through the system quickly.

Yes, I agree with that. And I just want to add that it is kind of a circular problem we have in measurement science, which is until you can show the effect -- you can show that your measure is altered by the drug, it's -- the uptake is just going to be limited. But as soon as a signal is detected, my other clients will come out of the closet, and I think it will become widely adopted.

So our next question comes from George Farmer at Scotiabank.

Really interesting presentation. Dr. Jaeger, I was wondering, could you comment on how this -- how EMACC compares to the Montreal Cognitive Assessment tool? It feels very similar in some ways, but maybe it's just that tool on steroids, it feels like to me. And then, are there any other tools that are being evaluated right now? And then also, how do we get an FDA advisory committee to kind of weigh in on all these different approaches and recognize that this is really a significant problem?

Yes. So MoCA, the Montreal Assessment, is a very brief assessment, 10 minutes or less. It is a better version of a mental status exam. And -- but it's still crude. It still doesn't have the -- so it doesn't have the refinement and the fully quantitative ability that we have here. It also has ceiling effects because it's designed to detect abnormality. So you can't get better. So the items can be performed perfectly, in fact, will be performed perfectly by most healthy adults. You won't make any errors. So you won't be able to see better than that. And some of our patients, if they come with high resilience, high educated people who are very good at this kind of thing, you can have mild cognitive impairment and actually perform close to the ceiling on a measure like that. This will have more measurement fidelity. And your other question was what…

How to get this in front of the FDA?

Yes. So typically, if you go to the FDA, there are different kinds of meetings. You can have a Type C meeting where you kind of say, well, here's what we're planning to do, what do you think? Invariably, they come back with, well, show us some data. It looks interesting, show us data. I've done it so many times. The other opportunity is what's called the end of Phase II meeting. So if you have a signal, you then take it to the end of Phase II meeting and plan with them your pivotal trial. So that is the decisive one. And obviously, at that point, the FDA can provide suggestions for alteration and so on if they wish. If they want a trial that's 3x as long and 10x as large, we can use the old measures. But I don't think that will happen.

Okay. Great. And then one for the company. You're also measuring CDR as well as EMACC in the study, right? Do you -- how are these tools administered during site visits by the patients? So I would imagine maybe they become fatigued having to do -- going through so many steps. Is that an issue? And how are you managing that?

Judy, I'm going to let you take this because I know that you design.

All right. So the CDR is a clinical interview. And as Sarah pointed out, you're spending a lot of time with the informant. So that mitigates fatigue while you're getting your blood draws and doing other things with the patient. Then you bring the patient in. And in a known patient, it's not that long, and it's sort of a chat with a little bit of cognitive testing and probing going on. And then the EMACC takes between 25 and 35 minutes to administer in total. We do worry about fatigue. We try to make sure that the testing is done at the same time of the day, each time. But of course, these factors are there in all the validation data. And in that 0.93 correlation, those people already had done the CDR and the blood draws and everything else is performing really well.

Yes. I would point out…

I would like to mention one more thing, if you don't mind, that both the patients who are receiving investigational medication and the patients who are receiving placebo are undergoing the same protocol, and they have the same opportunity for fatigue. So we would still be able to see a difference even if fatigue is impacting all the patients who are participating in the trial.

Yes. Okay. I'd also point out, if you don't mind, some of the docs we talked to say that the informant can also be cognitively impaired depending upon how old they are. So that certainly throws in another wrinkle.

Yes, we typically don't examine them. And so absolutely. That is a complete -- that is absolutely an issue when it's the spouse. In some ways, they're the best informant because they can see the behaviors, but yes, often -- or they're spouses. They have relationships. Maybe they disagree on something and they're calling it a memory impairment. So that's where the noise comes in.

Our next question comes from Joel Beatty at Baird.

Maybe first one is, could you talk us through the powering assumptions for AD02 and what helps get you comfortable in the 6-month 201-patient trial?

Want me to do that, C.J.?

Yes, go for it. You're on a roll.

All right. We used ADNI. We looked at the EMACC measures in the ADNI cohort. We looked at one-year change. ADNE is the Alzheimer's disease neuroimaging initiative. It is a longstanding U.S. study. These patients are part of multiple sites. It's a multisite natural aging study. All of these patients had confirmed mild cognitive impairment due to Alzheimer's confirmed by PET. And we looked at the slope change over one year, and we also then enriched it and looked at the slope change only in those who had biomarkers for inflammation. And by the way, the power doubled when we did that, just a point. We powered to the CDR change, the CDR Sum of Boxes change over 1 year, but we also calculated what the power would be using the EMACC. And so the study was powered for CDR, but it looked like we -- our sample size requirements for EMACC were about half those for the CDR. Those are data that we presented publicly at a previous CTAD meeting, I think 2021, we could share that.

Awesome. And then how often are these cognitive assessments like EMACC and other measures being looked at in AD02? Is it just kind of a baseline and at the end or other time points along the way?

No, I believe in frequent measurement. We're doing them. Are we at 6 weeks? I can't even remember, C.J.

Every 6 weeks for the EMACC.

Every 6 weeks, right? So yes, every 6 weeks, they come in and get the battery with the alternate forms. Obviously, they're not getting the same word list. But yes, same battery.

Yes. I would emphasize, this is a key point, Joel, that the frequency of measure improves the precision of the outcome, decreases the variance and gives you more confidence in the results.

And increases the power.

Everybody that's doing just a test at the beginning and the test at the end, they are really wishing on a star, so.

And maybe last question is when you did the interim analysis earlier this year, I know you looked at EMACC. Did you also look at some of these other cognitive measures and get any takeaways from those?

Yes, sure. We did. It was descriptive statistics. We don't have test retest on any of them, so we can't look at the agreement of the CDR administered twice. But the distributions are as expected. Obviously, the inclusion criteria included CDR-Global. So by definition, they're all 0.5 or 1. And the distribution of the CDR Sum of Boxes was appropriate, suited and fitted the trial. We had no other method for validation.

Our next question comes from Tom Shrader at BTIG.

Thanks for the presentation. It's an interesting approach. The -- C.J. took a little bit of a shot at this, but Judy, maybe you can comment on elements of this test that are inherently faster beyond the sensitivity. Do you agree with that? Did you see that as you looked at past trials that certain pieces of this test changed very rapidly? And is that something that in this trial, in this small trial, might give you almost equivalent to dose response, some piece you expect to change fast does change fast?

Well, fast is really effect size over time. And if you have a more precise measure and if you don't mind some statistics, the effect size is the mean change over the standard deviation. So if you reduce the variance and you measure more often, you see a tighter steeper decline, and it's not because it's faster. It's not picking it up faster, but it's more precise. So you have better signal-to-noise ratio. I mean decline is biologics.

So it's really just sensitivity.

Exactly. The decline is decline. So you'll pick it up faster only because you're going to pick it up. If you have an insensitive measure, you need more time to see it.

And then you said something that the INmune trial is powered for CDR. But aren't all the trials powered for CDR and they come up with at least twice as long and about 8x as big? So maybe you can dig in a little bit as to what you meant by that?

Do you want me to take this, Judy?

Yes, go ahead.

So this is the power of information, right? So information -- patients have inflammation irrespective of the metric. They have faster progressing disease. Now what I will say, I think this is powered off of sample of ADNI. So this is -- and to the extent that that reflects what we get in our trial, right, the broader population, it should be fine. But the issue is, as Sarah and Judy talked about, is that there's all these elements of variance that you can't account for. So to make sure that you're capturing those patients and you're accounting for that variance, typically they're larger and longer trials. But again, to the extent that the sample that we use to power the study is what we have in our trial, it should be fine. What's interesting is that the variance in -- that we saw in the EMACC is consistent with what we saw in ADNI, which is what you would expect to see from a tool that is objective, performance-based and really explains decline in the patient population. Judy, do you want to add?

Yes. No. I mean, I think companies power using different cohorts and data sets as well. So I don't have intimate knowledge as to how companies are using CDR to power and what cohorts they're using. So I can't be really precise in my response to that. But I do know that when we didn't restrict to the inflamed, I think we had more than double, almost 3x the sample size requirement for this study.

And Tom, I would add that if you look at the data, the Phase III data from lecanemab and donanemab, and you look at those figures on CDR, in fact they could have stopped those trials much sooner than they did and had statistically positive results.

That's the point I was just going to make. That trial in retrospect was incredibly overpowered.

Yes, I mean -- so whatever they were using, as Judy implied, they were using something that -- I don't know what they were using, but they really went for overkill. And as you know, that overkill probably cost them $100 million.

Our next question comes from Jim Molloy at AGP.

A couple of questions I have are, when will you know if EMACC is going to be acceptable to the FDA? Do you anticipate it will be after the end of Phase II? Or do you think it will be not until sort of you at an expected AdCom following NDA filing? And what are the 3 -- I think you mentioned earlier there are 3 companies currently running. Is that yourself and Biogen? Or what are the 3 companies currently running trial?

I'm afraid I have confidentiality agreements with them, and I can't share.

Okay. Is it -- because I know there's Biogen and you guys. Is there a third company or are there 3?

There is a third company, yes. Sure. Yes, I think realistically, end of Phase II meeting is -- will be the best shot.

Yes. I mean, just to be clear, we would not pick a primary endpoint for a Phase III trial without complete blessing from the FDA. I mean we're not -- even if -- as you've heard me say before, even if they make us do CDR, we can do that. We just know we're going to have to do a longer, bigger trial. We strongly believe that at the end of the day, given the right data, you've heard the background today when we provide a validating data set, the FDA wants to move this field forward, number one. Number 2, they don't want to put patients at risk. Remember, when you take a patient and put them into an 18-month trial on placebo, you take someone who might have MCI. But 18 months later, they may have advanced on placebo so dramatically that they are no longer eligible for any kind of therapy. So it really becomes a bit of an ethical and safety issue for the FDA. They want to eliminate that problem. And one of the ways to do that is smarter, shorter, smaller trials. And if we provide the tool, data-driven as Judy likes to say, we believe that they will take advantage of it.

Great. And maybe a quick follow-up then. R.J. or C.J., what sort of feedback, pushback did you guys get last week's CTAD in Madrid? Any anecdotal stories from the community talking about EMACC there?

Judy?

Yes. I actually presented that poster really, really favorable. I would not be at all surprised if we don't see another company or 2 pick it up. I mean correlation of 0.93 on cognitive tests is truly striking. Yes. No, those who stopped and looked extremely favorable.

Yes. I think just to add a little bit, there's a little bit of a dissociation between neuropsych clinicians that really understand measurement of cognition and companies who sort of don't really know anything about it, but bring these people on. And so you don't hear a lot about from the companies per se because they're really not in the details. But when you talk to the neuropsychs and every time I'm around Judy and her colleagues and some of these other folks that really understand this stuff, they go, yes, it's a no-brainer. I mean it's very, very clear. And so now we've got the data, the validation data to move it forward. And I think it's going to be just incredibly well received. And I think other companies are going to go, "Hey, thanks for doing this."

So our final question comes from Jason McCarthy at Maxim.

Sure. A lot of questions, and it kind of -- it's a good segue into me taking it in a little bit of a different direction to finish it up. So obviously, all incredibly innovative, good stuff. You had mentioned, R.J., maybe somebody had typed in a question online about Wall Street being nervous. There's more risks, smaller trials, faster trials in Alzheimer's disease. So my question is, who has to buy in first to this, right? Wall Street sometimes doesn't like change, right? Is it going to be the juggernaut of amyloid with big pharma and the tens of billions they spend, do they have to buy in first? Is it investors who have to start to buy into this? Or obviously, the FDA does, but just your high-level thoughts on where this goes.

Yes. Let me reinforce what Judy just. First of all, you're absolutely right. The FDA is the top dog here. I mean until the FDA finally makes the comment and blesses it, it will -- you on Wall Street will always worry. I mean that's what you're paid to do, right? But remember what Judy just said, she said that people came to our poster. They were really interested, and she predicts that people are going to pick it up because I can tell you, if you're a small company, you can't do -- it's difficult to finance a 500-patient trial. It's difficult to finance a 200-patient trial, 201-patient trial like we have done, but at least it's attainable. And so back to Judy's point, if we are going -- we, meaning both biopharma, the investor community and the clinical community, are going to expand the offerings to Alzheimer's patients, we can't do it with 18-month 1,600-patient trials. In fact, we can only do it by using smaller, smarter, shorter clinical trials. And to do that, you need a tool like this. And C.J. said it, the dam will break. The first -- when we go there, when we ring the bell and go there and tell the FDA, look, here's the result, blah, blah, blah, they give us the thumbs up that literally the dam will break. CDR will be out of business in this field. That's my prediction.

Good enough.

So with that, this has been really, really interesting, guys. I hope you've learned as much as I have through this process. There are extra questions that have not been answered on email, which we will get to. We promise to answer them. But thank you for your time. I won't apologize for going over because I think all of this time has been extremely well spent. So with that, Tara, to you.

Thank you, R.J. So this concludes today's event. You may now disconnect.