Hey, this is Mani Foroohar, senior analyst, generic medicines here at Leerink Partners. Doing a quick intro here for the team from Passage Bio, for a presentation this morning at the 25th Annual Leerink Partners Conference here in Miami. I'm going to pass the mic over. Thanks, Mani. All right. Hello. My name is Will Chou. I'm the CEO of Passage Bio. I'm going to be talking today primarily about our lead program, PBFT02. It is a gene therapy with the lead indication that we are in the clinic for is FTD with a granulin mutation. I'm going to spend some time today talking a little bit about FTD, the pathogenesis of FTD- GRN, a little bit about our preclinical data, a lot about our emerging clinical data, and then finally a little bit on regulatory and manufacturing. What is FTD- GRN? Frontotemporal dementia is one of the most common early forms of dementia. FTD- GRN is a genetic form of FTD. It's caused by mutations in the granulin gene, which leads to a low level of progranulin. There are no approved disease-modifying therapies. PBFT02 appears to be a differentiated, potentially best-in-class product. It is a one-time AAV1 gene therapy that is administered directly to the CSF. Our early clinical data shows durable, elevated CSF levels of progranulin. The market opportunity is quite significant for FTD- GRN. This is an orphan disease, but not an ultra-orphan disease. There are about 18,000 FTD- GRN patients across the U.S. and Europe. We are also beginning to study PBFT02 in the FTD- C9orf72 population. There are 21,000 of those patients across the U.S. and Europe. There is evidence that raising progranulin above normal levels can address the underlying cellular pathology in FTD- C9, which is TDP-43 pathology. We have reached agreement with the FDA to open a study in ALS, also using PBFT02. We are currently undergoing some preclinical work on raising progranulin in Alzheimer's patients who have a predilection to have lower progranulin levels, those with the granulin SNP. I am going to come back to milestones at the end of the presentation. Let me tell you a little bit about FTD. As I mentioned before, it is one of the most common early-onset dementias. On average, patients live about eight years after the diagnosis. The important thing to mention is, one, there are no treatments for it. Two, it is a behavioral dementia. You can see at the bottom of the slide with the little icon, some of the bizarre behaviors that most patients exhibit. It is important because it is difficult for family members to bring these patients out into the world, such as going to a doctor's appointment. When you think about differentiation with other therapies that are in the market, a one-time therapy has a huge unmet need, especially when compared to a therapy where one would have to go take your loved one who has a behavioral dementia once a month to the hospital. In FTD- GRN, the low level of progranulin is the proximal cause of the disease. Progranulin is necessary for normal lysosomal function. With low levels of progranulin, you get lysosomal dysfunction, pathogenic inflammation, and TDP-43 pathology that ultimately leads to neurodegeneration. The interesting thing about the progranulin space in FTD- GRN is because of the normal homeostasis of progranulin, this vector does not need to transduce every affected or at-risk cell. Because of the normal homeostasis of progranulin, we can benefit from what we call cross-correction. Progranulin in its normal state is excreted into the extracellular fluid. It is then endocytosed into neurons and microglia via a variety of cell membrane receptors. Once inside the cell, it exerts its positive function on lysosomes. By increasing the amount of progranulin in the extracellular space, we increase the amount of substrate that is available to be endocytosed into the cell. That is why we can leverage cross-correction. We can transduce all kinds of cells in the CNS. The cells you want to transduce are everlasting cells. You do not want to transduce cells that die off because then you lose your signal over time. You lose your progranulin level. We have actually seen that in some other products. Okay, why did we choose an AAV1 capsid? This is the non-human primate study that really drove that. What you see here is human progranulin levels on the Y-axis. The big spike line is an AAV1. The other lower lines are AAV5 and a modified AAV9, which is labeled here as AAVhu68. You can see that AAV1 had a dramatically higher level of progranulin production in non-human primates. Honestly, when I share with you the clinical data, we've seen the same effect in the clinic. The one other piece of preclinical data I'll share here is a distribution study also in non-human primates. There has been some discussion that the ICM, so intracisterna magna route of administration, we've heard some say that it does not result in distribution of vector throughout the brain. That is not what we found in our non-human primate studies. This is an ICM injection of our vector into non-human primates. In every part of the brain that was sampled, cortical and subcortical, vector was found in a dose-dependent manner. Now on to our clinical study. We're in the midst of a phase one-two study. It is a multi-center, multinational study. We have seven sites open across the U.S., Canada, Brazil, and Portugal. We'll be opening a few more sites this year in a couple of new countries. It was originally intended to be a dose escalation study with three different cohorts, with the opportunity to escalate dose. Because of the very robust progranulin response we've seen at the lowest dose, we're actually moving down in dose in the midst of the study. We've dosed seven patients. We are in the middle of cohort two. We are moving now down from a higher dose, dose one, to half that dose, which will be dose two. The total genome copies that we're administering are 4E13 for dose one and 2E13 for dose two. This is a very, relatively small absolute amount of genome copies. Intracisterna magna, what is this administration? This is a brief, it takes about 45 minutes, CT-guided administration directly to a pocket of CSF at the base of the skull that is outside of the brain parenchyma. The needle never penetrates any brain parenchyma. There are no concerns of micro-lesions to the brain. This procedure is done by an interventional radiologist. It does not require a neurosurgeon. A lot of availability issues with getting a neurosurgeon, there are a lot more interventional radiologists. By delivering directly to the CSF, we get a broad CNS biodistribution, which we already went through with the non-human primate work, lower doses, which I've already talked about, and also the reduced impact of neutralizing antibodies and the potential for reduced immune response. Because of this last point, when we started out this study, we started, the trial started with a very low level of immunosuppression. The starting immunosuppression was just 60 milligrams of oral prednisone. I remember when I started at Passage Bio, I asked, why are we starting so low? The answer was, it's because this is a protected space. The intention is that there would be a much lower immune response. As it turns out, there's still an immune response. Sixty milligrams of oral prednisone was probably too low. It was definitely too low. This is the tox profile of the product. We've dosed seven patients. In five patients, there have been no SAEs, so just mild treatment-emergent AEs. Two of the patients have experienced a total of three SAEs. The first patient, remember, they only got 60 milligrams of oral prednisone. They had two SAEs. They had hepatotoxicity. They had a big bump up in LFTs that responded to IV steroids. They also had a venous sinus thrombosis that was asymptomatic. After that patient, we increased the immunosuppression. Now we moved all the subsequent patients, have received a regimen of three days of IV methylprednisolone, one gram of IV methylprednisolone for three days, followed by 57 days of oral prednisone, just 60 milligrams of oral prednisone. Still just a steroid regimen, but that appears to have taken care of any systemic immune response. No other patients have had any other LFT abnormalities. One patient, patient seven, also had an asymptomatic venous sinus thrombus that completely resolved with anticoagulation and remained asymptomatic throughout. This patient is still in the study and we are continuing to follow them up. There have been no complications from the ICM administration and no DRG toxicity. As I mentioned before, the remaining patients in our second cohort and likely in the third cohort also are going to receive dose one, which is 50% of dose two. Our hypothesis for the VST, because now there are two patients who have had a VST, is we could be seeing a local inflammatory response with the transduction that is causing a hypercoagulable state. By lowering the dose, we are going to lower the inflammatory response as well. The goal is to have a more favorable tox profile without having to add something like prophylactic anticoagulants, which we absolutely could do, but we want to give the chance for patients to not have to have that with a lower dose first. The reason that we can move to a lower dose is because of the really incredible target engagement that this product has showed with just the lowest dose, with dose one, or with the initial dose, dose one. What you see here on the Y-axis is CSF progranulin levels in our first six patients who we've treated. The gray area at the bottom is the normal range. That's about 3-8 nanograms per mL. The average is about 4-5 in a normal patient. There is a, just for relative speaking, there is an anti-sortilin antibody that is administered once a month. It gets to levels of about four to five, so right in the middle of that gray area. There's also another AAV9 that is administered ICM. It peaks at about two months at about levels of 14-15, and then by 12 months has trailed down to levels of about six to eight. We are really pleased with the profile of this product to date. All of the patients by six months are well above the normal level. The product has shown durability out beyond 12 months. We have two patients durable at 12 months, and the longest patient, this product's been durable out to 18 months. There are no signs of any waning of the progranulin effect. Now, we are asked, I'm asked all the time, how do you know that higher progranulin actually matters and that it will make a difference to patients? The answer is, we do not know yet. Nobody knows yet. That is why we are doing the study. Nobody knows what the threshold is for how high one needs to get CSF progranulin in order to have a clinical effect. In fact, it could be different for every patient depending on where they start in their disease state. Because we do not know what the threshold is, if on average we can get to higher levels of progranulin, then you can see the dispersion. Any product, there is going to be a dispersion of higher and lower levels of progranulin. If you can get on average to higher levels, your probability of actually being above that threshold is higher. This is just plain old drug development. Higher target engagement in general is better. Plasma progranulin levels should not go up, and they do not go up. These patients start at a low level because they are haploinsufficient in the GRN gene, and there is really no effect on systemic progranulin. We are trying to drive progranulin levels where it matters, which is in the CSF. Some data that we shared just earlier this year is on plasma neurofilaments. Plasma neurofilaments are a sign of neurodegeneration. This is the only biomarker in the FTD space where there is published natural history. We actually know from abundant natural history data the course, the normal course of plasma neurofilaments in untreated patients. In symptomatic untreated FTD- GRN patients, the annual rate of change is about 29% a year. For our two patients who have reached 12 months, this is very early data. It is only an N of 2, but it looks very promising. On average, they're at negative 13%, which is a marked difference from the expected annual increase in neurofilaments. It does appear that the product is having an effect on neurodegeneration. This is a great leading indicator for us. Finally, just to recap the competitive space. PBFT02, it's an AAV1. We're in a phase one-two. It's delivered ICM. It's a one-time therapy. The product has been able to achieve levels of 13-27 nanograms per mL of CSF progranulin and has shown durability out to 18 months. There is also an anti-sortillin antibody that is reading out phase three data at the end of this year. It's administered monthly. It gets to levels of about 4-5 nanograms per mL, and we don't really talk about durability because it's a monthly product. I can, I do want to come back to the clinical features of FTD now. Remember, this is a one-time therapy versus a monthly therapy. For patients who have a behavioral dementia where it's difficult to go out into public, we do think even if there's equivalent efficacy, there is a place for a one-time therapy to avoid the multiple trips to the hospital that one would need pretty much forever with a monthly therapy. Finally, there's an AAV9 therapy that's also in a phase one-two. It's delivered the same way, ICM. At 12 months, the levels are about 4-8 nanograms per mL, so much lower. The pattern of CSF progranulin is a continuous decline from months two to twelve. Finally, looking ahead, first, one of the, a couple of the big hiccups that gene therapies have had as they near commercialization have been in manufacturing. One is aligning on a functional potency assay with the agency. We have done that. We developed in-house our own functional potency assay, and we have aligned with the FDA that this is fit for purpose. The second is scaling up an academic manufacturing process into a commercial-ready process. We just completed development of a high-productivity suspension-based process. We finished a 200-liter run with excellent results from a quality standpoint. In terms of productivity, this 200-liter productivity, we could treat about 1,000 dose two patients with a single 200-liter batch. The efficiency is excellent of the new process. The other thing on registrational studies is they're, we're going to seek feedback in the first half of 2026 on registrational strategy. There have been some recent gene therapy precedents in adult neurodegenerative disease utilizing natural history data as an external control. It's a huge boon for the field. It's encouraging, particularly in FTD- GRN, because similar to other diseases such as Huntington's disease, there are existing natural history databases that are well-organized and prospective of FTD- GRN patients. Finally, our milestones ahead. As I mentioned, we will seek regulatory feedback on our registrational trial design in the first half of 2026. In the second half of 2025, we will report additional 12-month data from dose one and interim data from the lower dose, dose two. We should be initiating dosing of FTD- C9 patients shortly in the first half of this year. Also, in our pipeline, we have a preclinical Huntington's program. We have $77 million in cash as of the end of the year and runway into the first quarter of 2027, which gives us plenty of time to watch this data mature and continue to treat patients. With that, I want to thank you for your attention early in the morning and look forward to any questions. A quick question. Sure. I guess more broadly, there's been a lot of debates on changes in regulatory infrastructure and the FDA's attitude towards genetic medicines, gene therapy, gene editing, we'll say therapies with a therapy of curative intent, broadly speaking. That was talked about a little bit at our panel yesterday with Alex Azar. Talk to us a little bit about where you are in terms of how you think about for FTD- C9 or patients. What's the ultimate goal you're going, you're looking at for potentially registrational endpoints? Do you see any change in the regulatory infrastructure, in the regulatory attitudes, and feedback you're having at the FDA? I know it's relatively early. Yeah. Where is the, where are you in terms of what you're seeing in changes in the environment or speed and responsiveness from the agency? Yeah, recognizing we're in a disruptive time. Yeah. I can just, I can't predict the future, right? There's been obviously a lot of changes, but for what we've seen for the last year, we had three regulatory interactions and they were very fast responses and all productive. In fact, so productive that we canceled all three meetings and we just accepted the response. There was really nothing to debate. From our experience, it has been excellent. What I'm hearing broader also is that particularly around, so manufacturing, there is more understanding and desire to work with sponsors to make things happen. It is very different in my mind from when we started. I started with Kymriah a while ago, and that was a very different attitude because it was early on. I definitely feel it from a manufacturing standpoint. From a clinical standpoint, I mean, we gave the regulatory example. This is obviously uniQure. They had some really excellent news. This is adult neurodegenerative disease. They will be there. They announced they're going to be filing a BLA based on a comparison to natural history, which is fantastic for patients and shows really some latitude from the agency because there's no good therapies right now. I think it's very encouraging. I cannot predict anything in terms of bandwidth of the FDA, although I just heard yesterday that it seems like bandwidth is going to be good, that there are no reviewers are part of any downsizing. I would hope for that. I know obviously you were, this is actually probably more relevant to the next slide. Okay. We talked about the data you're having later on this year to trigger regulatory interactions. Yeah. How do you think about the goal posts for a registrational study? Like what's, you obviously can't, without the regulatory feedback, you can't say, oh, that we think our pivot will be X. Yeah. I get that. but like what are the goal posts based upon your ongoing interactions in terms of what a potential endpoint could be, study size, et cetera? Like what should we be benchmarking against? Yeah. I think there are, there's a couple of spectrums of outcome that I would think about. There is the far end of the spectrum, which is the traditional randomized placebo-controlled study. It's large, takes a while to do. You need a lot of sites. The other end of it would be a comparison to natural history, which we've also seen, much smaller but still robust if you have a good enough natural history database. I would say the question in there is particularly around FTD- GRN where neurofilaments progress and there's good natural history of the progression of neurofilaments. We saw it in the data that we shared. This is already, that's already been published and there's a lot of natural history. The question in my mind would be, how good of a biomarker is this for an accelerated approval, knowing that we have a good clinical endpoint as well in CDR? Operationally, I think a question that I've had with another in this room yesterday with another gene therapy company in a very different patient population, but with a similarly complex delivery and complex as in not IV and not subcu. Sure. Talk to me about blinding in a potential study, and the relevance of how well would you control arm, is there a sham procedure, you know, the extent to which that might be practical and the extent to which there's an unblinding risk because this is not, it's not a completely non-invasive approach. Sure, sure. You raised very important questions that we want to talk to the agency about. Yes, you can do a sham placebo. You have to anesthetize and sedate a patient. You have to give them a prick. You have to do some sham steroids as well. It is quite involved. There are some real complexities to it that maybe are not as good for a randomized placebo-controlled study. I think you raised very important points that we intend to speak to FDA and EMA about. It is doable, but it is quite complex. I know you have not had those conversations yet. Can you give us a sense of where these patients are in terms of their function and ability to sort of that kind of invasive sham or we will say that kind of thorough sham? It can be quite burdensome for patients that have more limited function. You start to get into sort of caregiver burden issues, recognizing that is kind of what you are trying to create a placebo for. I agree with you wholeheartedly that this is, I was a geriatrician before. This is patients who have dementia of different levels. When you give them something that affects their mind, you put them at risk of something like a delirium, right? Sometimes when you have a delirium, you just ratchet down from that. If you're offering the benefit of a one-time therapy that has potential to really make an effect on their disease state, then yes, that may be worth it. Just for a placebo, there is a question of is that ethical to do that? I think it's a reasonable question. I know I'm eating up on the same question. For patients in the population who are going to be enrolling, what is the incremental increase fall risk for putting that patient going through the whole process of anesthesia and out of it? I can't answer that question with a specific number, but you must have read my old research as a geriatrician. I was on fall risk. It's absolutely higher. Once you put a patient through that, you absolutely increase their fall risk. In terms of bringing this to an IRB and talking about a full sedation for placebo, I think it's going to be difficult. That is what we want to talk with the agency about. I think we're all going to be watching very closely because these are questions across a lot of these patients, across age ranges who are a little more, whoever developmental disability at one end of the spectrum or dementia delirium at the other end. It's going to be an exciting time for you guys throughout this year data-wise and an unpredictable time in D.C. as always. I'm looking forward to continuing this conversation throughout the year and seeing more data as the year progresses. All right. Excellent. Thanks, Mani. Great questions. All right. Thank you.