Joining Guggenheim's 2025 Healthcare Innovations Conference. I am Debjit, one of the therapeutic analysts, and my privilege to welcome our next presenting company, Passage Bio, and joining from Passage is President and CEO William Chou. Good morning. Thanks for having me. Thanks for making time for us, and, you know, as is customary, let's do a quick intro before we get into the Q&A. Sure. Will Chou. I'm the CEO and Acting Chief Medical Officer of Passage Bio. We are a gene therapy company. We have a lead clinical program in frontotemporal dementia with the granulin mutation, and we also have a preclinical program in Huntington's disease, so look forward. I think we're going to be talking mostly about the FTD program today. Yeah, I mean, that's where the interest is. That's where the interest is. You know, you've been working on this for a good part of two years now. Exactly. In the clinic. Yeah. So why don't we start off with the recent phase III disappointment from one of your competitive programs? Sure. What does it mean for CSF PGRN levels? Sure, so we're certainly very disappointed for the FTD-GRN patients. This was a big hope for them. I think about progranulin levels. You know, we thought about the different hypotheses for why that study may have failed. Progranulin levels is one of them, so from that, latozinemab's phase II data, the average CSF progranulin level was about four to five nanograms per mL. The normal range is three to eight, so they're able to get patients from two to three to about four to five. Now, if the average is four to five, that means some patients may have been getting in the low threes, and perhaps that wasn't high enough, so because nobody knows what the threshold of how high progranulin needs to be, perhaps the product just didn't raise levels high enough. That's one of the hypotheses. There was the second blocking sortilin. What is it to? What are you doing differently? Sure. Mechanistically, the way that product works is progranulin in its natural state is an extracellular protein that is endocytosed by a variety of cell membrane receptors, and it exerts its positive function intracellularly on the lysosome. It has to get inside the cells. Now, by blocking sortilin, latozinemab decreased the destruction of progranulin, but also might have inhibited the intracellular function of progranulin, the positive function on the lysosome. The mechanism may have actually potentially put a cap on the clinical efficacy that one might see. That product was actually disrupting a natural mechanism in the cell, which is sortilin. The AAV approach, and we are one of several programs using AAVs, are actually generating more progranulin. None of the AAV programs are actually disrupting any of the normal processes in the cell. We are creating more progranulin in the patient, and that creates more substrate to be endocytosed intracellularly, which is where you want the progranulin to be. So if you combine the lack of meaningful increase in progranulin and the impact on sortilin, the two of them together prevents or impacted the impact of the readout on neurodegeneration. I would say those are potential, those are hypotheses for why it may not have worked. None of us know what actually happened, but those are two very leading hypotheses, level and actual mechanism. So from your perspective, meaningfully increasing progranulin levels, that's still an active product or things that you really want to investigate? Absolutely. So the Alector study disproved a hypothesis. It disproved the hypothesis that raising extracellular progranulin by blocking sortilin is going to help patients with FTD-GRN. And that's a very specific hypothesis. It did not disprove the hypothesis that if you raise progranulin levels in a different manner, such as an AAV, that that won't affect the clinical outcomes of patients with FTD-GRN. This is in the end a neurodegenerative disease? Correct. Could you talk to the baseline of the Alector program and what you are doing differently so you can tackle the disease early on? Sure. Absolutely. So this gets at the heart of our third potential hypothesis for why that study may have failed. The way the Clinical Dementia Rating Scale for FTD works is there is a Global score. And that Global score goes from zero, which is unaffected, to three, which is the most severely affected. Now, the Alector study and our study previously included patients who were at a level of one and two. The Alector study capped the number of two patients at 30% of the total study. And that was for a good reason. If you have too many patients who are too severe, your likelihood of actually seeing a clinical response is lower. As you mentioned, this is a neurodegenerative disease. To actually see a clinical response, you need to catch patients as early as possible. Once patients have declined so much, none of our therapies are replacing neurons or making them grow more neurons. They're all preserving as much as possible. We're trying to interfere with the disease mechanism to preserve, so you don't want patients who've slid too far, so the Alector study capped the 2s, the CDR 2s, the more severe patient at 30%. We had made the decision even before hearing about the Alector results that for our cohort 3 and moving forward, we are going to exclude CDR 2 patients from our study and only include patients who are CDR 1 or 0.5. 0.5 is equivalent to mild cognitive impairment in Alzheimer's dementia. These patients have deficits, but they don't completely qualify for FTD yet, but they have the mutation. You have to have the mutation to be in our study. We know 100% of patients who have the mutation will get the disease and are going to progress. We've already made the change in our study to target an earlier population of patients. I would say that if you have the wrong patient population, you could have a great product with a great mechanism in neurodegenerative disease and not see clinical results. So if you were to take this program to a registration study. Limiting 0.5 to one CDR. How quickly can you enroll a study and what kind of, or how many patients are available? Or how quickly is the disease diagnosed to capture enough of these patients? Absolutely. So the epidemiology of FTD-GRN is in the U.S. There's about three to six thousand patients based on epidemiologic projections. We know we need genetic testing, though. So patients need to be genetically tested, and they need to be tested as early as possible. We also know from looking at many other disease areas, if there is an actual therapy on the market, patients and people at risk, family members at risk, are much more likely to get genetically tested. What is the reason people don't get tested now? They ask, what's why? Why should I do it? Why would I want to know? There's nothing I can do about it. It's just going to make me more anxious, right? That is a big concern, and it's very valid. But if there is a therapy, patients will get tested earlier, and we definitely want to catch these patients early. I would imagine any living relative of somebody who is diagnosed, once there is an actual therapy available for FTD-GRN, almost all of these patients are going to want to get tested because you're going to want to intervene early. Do you recall how long it took Alector to enroll the study and how? It took them a few years to enroll the study, definitely. So now if you eliminate 30% of the patients, does that impact your planning going forward? We eliminated the twos, but we also added the 0.5s. That's one. Second, I'll say there is much more; there are multiple trials ongoing. Now, in one sense, that's competitive, but in another sense, there is a lot of excitement among patients about getting tested and getting involved in studies. So as the patient community hears more about it, they get more engaged. And I will be honest, we haven't had any issues with recruitment and getting patients involved in the study. And given the disappointment for Alector, does this limit you to conducting a randomized study before FDA actually entertains an approval? Or given that this is a gene therapy modality, you have clear mechanistic differentiation, there could be some degree of flexibility? Based on what we've seen from recent guidances from the FDA, such as the rare disease evidence principles that were just published in September, we do think there is more of an openness to single-arm studies under specific conditions. So what the rare disease evidence principles stated was if there is a disease that is relatively low in prevalence, it progresses very fast, you have a mechanism that addresses the genetic cause, and there's no treatment available, they would entertain single-arm studies. So it's something that has always been understood that these were the requirements, but we find it very encouraging that they were put on paper, they were signed off by FDA leadership and published. And that is putting a stake in the ground for how they take these things. So we're encouraged by that. And we believe that FTD-GRN does fit into this criteria. Got it. Given that you had to walk down the dose a little bit, can you get or can you maintain the GRN levels above the 10 nanogram threshold? I think it was nanograms, right? Yes. Nanograms per mL. Absolutely. Yes. So the normal range is three to eight. And we've only released some very early data from our lower dose, dose two, but at 30 days, CSF progranulin is already near eight. And looking from our dose one data, we know that the CSF progranulin levels continue to go up for at least three to six months. So far, it looks very encouraging. Our goal is to get CSF progranulin levels at or above that upper limit of normal. So if we get average levels at or above that eight, that would mean that even the people who are below average are still in that high range. And so that's our goal. Right now, it looks like dose two, at least with an N of one, is on track to hit that goal. Got it. What's the fundamental grounding philosophy that increasing progranulin levels will actually stop the progression of the disease? Given that Alector sort of, despite the limitations of that study, there are clear questions around this right now. About the mechanism. Progranulin, the progranulin deficit is clearly the driving proximal cause of the neurodegeneration in patients with FTD-GRN. There is no doubt about that at all. They have low levels of progranulin. Patients with FTD-GRN have levels that are around two, much lower than the average, which was about five, and mechanistically, it's very clear that that is the proximal cause for why these patients have neurodegeneration. As I mentioned before, blocking sortilin increases extracellular progranulin. You can't actually measure intracellular progranulin, so it's not clear that the progranulin was actually getting inside the cell where you need it to be, so I don't think anything has been disproven at all about the benefits of getting progranulin inside a cell. If you're just measuring extracellular progranulin and you're getting that by blocking sortilin and not actually creating any more progranulin, I don't think we've disproven anything about the hypothesis that replacing progranulin can have an effect. Help us understand how quickly the disease progresses, even when you have patients with relatively early, I mean, enrolled early in the disease progression? How much, what's the window of opportunity that you have? Sure, sure. So the disease progresses over the course of about eight years from diagnosis to death. Unfortunately, it's quite rapid. So catching patients early on is important. And it is relatively consistent over time in terms of the progression. As soon as patients are diagnosed and have that 0.5, they do start to progress relatively quickly. So we do have a window of about half the time to catch these patients. So you talked about the FDA guidance. We just went through this with uniQure, where the biggest pushback from the FDA apparently was you did not pre-specify your stats analysis plan when you started the study. The study is, you know, at best, signal finding. So, what are you doing to? If you can proceed with a single-arm study, what's the plan? Sure, so that was a disappointing outcome, of course, what we heard from uniQure. I will say that that criticism of not pre-specifying the stats plan before, I mean, I've been working with the FDA for a long time, and that is always the mantra. You need to pre-specify everything before you actually run the experiment. That is something that has always, there are made exceptions to it, but that is kind of a core tenet of what the FDA guides. So what we're doing is we're having an initial discussion as early as possible. So the FDA encourages early discussions on design, and we have guided that we will be doing that in the first half of next year to understand from the agency what the threshold of experimental design is that they are looking for in such a rare disease. We're going to have that conversation early so we can design based on their feedback. Got it. In terms of the immunosuppression protocol, there have been changes to that. How comfortable are you right now that every parameter is addressed? Sure. So the immunosuppression protocol, we changed it very early on in the study, really after the first patient. So since the first patient, it's been quite steady. We give three days of IV methylprednisolone, so pulse steroids, followed by oral prednisone. We have also added low-dose prophylactic apixaban, which is Eliquis. Eliquis is already approved for patients post-hip fracture or post-hip replacement and knee replacement at a low dose for DVT prophylaxis. And it's been proven to be very safe. So a major bleeding rate of less than 1% in these patients who are post-op. So we feel very comfortable, and all of our investigators feel very comfortable about this type of a short-term intervention for patients. So it's going to be several weeks of low-dose Eliquis as well as prednisone. How robust do you think natural history data are, and how much of that can you leverage? Sure. So the great thing about the FTD community is that there are two large, well-organized, prospective natural history studies, one based in North America and one in Europe. One is called ALLFTD. The other is called GENFI. They have many patients enrolled. They've been following patients, and they continue to follow patients longitudinally. They collect all the same parameters that we collect, including CDR, Global and Sum of B oxes, including plasma neurofilaments. They don't collect CSF neurofilaments because obviously it would be very difficult to take CSF from a patient in a natural history study. But those are two really important outcomes that there is a wealth of natural history data on. You had about a 2%-3.5% increase in NfL? Yes. It was very different from what we've seen in natural history. Yes. How confident are you that delta is going to hold up and it's not a function of small N? Sure. Well, that is the goal of the product, is to slow neurodegeneration. And this is the best biomarker that we have for is it slowing neurodegeneration or not. So there has been a wealth of previous evidence associating plasma neurofilament absolute levels with MRI changes, with cognitive function, and also changes in plasma neurofilaments with changes in brain volume and changes in cognitive function. So we think that it's a fairly good biomarker. And there is, as I said, a good amount of natural history evidence. I'll also mention, you said that we showed about a 3% increase in plasma neurofilaments. That is very close to the range of just normal age-related change in plasma neurofilaments. Unfortunately, all of us, we all go up by about 3% a year. And so reducing it to the natural rate, we think is pretty significant. Got it. In terms of prioritization of clinical programs. GRN, FTD, and C9. What drove you to the second program? Yeah. So in terms of prioritization, the vast amount of our resources are put on the FTD-GRN program, no question. The reason we've opened this study also to FTD-C9 patients is that there is some good preclinical data showing that raising progranulin levels can ameliorate TDP-43 toxicity. And we feel that FTD-C9 patients deserve that chance. And in a preclinical model, it's actually very hard to completely replicate everything that is happening in a patient. From a resource perspective, the biggest costs of running a study are setting up the infrastructure, setting up the sites, getting contracts with sites, setting up your CRO. The actual incremental cost of treating another patient is actually very low. So we already have a study with FTD set up. The additional cost of adding FTD-C9 patients is quite low compared to the whole study. Given the level of unmet need in FTD-C9 and given the preclinical evidence we've seen, we decided it was worth a try for patients with FTD-C9. Fast forward to phase I-B top-line next year. Give us a sense of the scope of the data that you plan to share? Are you planning to meet with the FDA before you share the data with the street or following? Yeah. So what we've guided to is first half of next year, we will do a data refresh. And we will also get guidance from the FDA on the registration path. I do think the guidance from the agency is a very important potential catalyst for us. I think given all that has happened recently, particularly you mentioned uniQure, there is a question mark out there, which is, okay, we believe in replacing progranulin. We believe that if you give enough time for this product and you follow patients long enough and you have the right patients, it should work. But there's an unknown about what it's going to take to actually get a product like this approved. And so we think taking some of that unknown off the table is going to be very important for us and for potential investors in us. And so we're really looking forward to getting that feedback in the first half of next year. Got it. So what I was trying to get at is when you release the data to the street, you also have the feedback from FDA at hand. So you get the data and we know what the path forward is. Or is it going to be sequence data first and then talk to the FDA, et cetera? You know, I actually don't know the answer to that yet. It depends on when the timing of everything happens. So truly, I don't know. But what I can say is that there's not a specific data signal that we are looking for to go have a conversation with the FDA. So the purpose of this conversation with the agency, this is not an end of phase II study where you come in with a set amount of data and you're, this is, that'll be later. We're not done our phase II study yet. We're still, we just started cohort three. This is a first discussion. We think it can be very substantive. And we expect it to be very substantive. But there's not a specific data package that we're waiting on for. Got it, and gene therapy manufacturing. We've got to ask, how much control do you have right now over the manufacturing? We work with Catalent. All of our product has been manufactured at Catalent. Our product in the future will be manufactured at Catalent. We have no need currently to manufacture any more supply. The next supply that we would be making would be for our registrational cohort. As we've discussed or as we released recently in earnings, we have developed a very high productivity suspension manufacturing process. One 200-liter batch can treat over 1,000 patients. We have a full capsid rate of greater than 70%. We have aligned with the agency on a path, on an analytical comparability path to pivot from the adherent process to the suspension process. Our goal would be to do that prior to starting the registrational cohort. Got it. And in terms of your cash balance right now, where does that get you? Sure. So our cash balance gets us into the first quarter of 2027. And we are spending about $30 million per year. So ample time for us to have these major catalysts that are happening in the first half of next year. Thank you all. Looking forward to the data on 2026 on the FDA guidance, and hopefully there's life on the other side. Thanks, Debjit. Thanks for having me. Take care.