Hello, everyone, and welcome to Oppenheimer's 36th Annual Life Science Conference. I'm Jay Olson, one of the biotech analysts here at Oppenheimer, and it's a pleasure to welcome you to our discussion with Passage Bio, and it's an honor to introduce William Chou, the President and CEO of Passage. Thank you so much for joining us here today, William, and with that, I'll turn it over to you. All right, thanks for having me here, Jay Olson. William Chou, I'm the President and CEO. I am a physician by training and have been with Passage Bio for about 3 years. Looking forward to telling you our story today. I'm going to spend most of our time today talking about our clinical program in frontotemporal dementia with the granulin mutation. I also want to mention we have a preclinical program in Huntington's disease. I'll touch briefly on that at the end. We have cash runway out into the first quarter of 2027. With catalysts that we expect this year, we expect 2026 to be an exciting year for Passage Bio. The lead clinical program that we have is called PBFT02. That is the name of the asset. It is to treat FTD with the granulin mutation. In frontotemporal dementia, there are no approved disease-modifying therapies, and we are treating a subset of patients who have a specific mutation in the granulin gene, and we'll talk about that in a second. Our product is a differentiated, potentially best-in-class AAV therapy. It is an AAV1. It is delivered via a nonsurgical injection directly to the CSF in a procedure that takes about an hour in an interventional radiology suite, and we've been able to show the highest and most durable levels of CSF progranulin of any product in this space. FTD with the granulin mutation is a rare disease, but it's certainly not ultra-rare. There are an estimated 18,000 patients with this across the U.S. and Europe. We are also studying PBFT02 in FTD with the C9orf72 mutation. There's a different rationale for FTD C9. I'll touch on that at the end of our presentation. As I mentioned, we have a preclinical program in Huntington's disease as well. Let's focus first on FTD with the granulin mutation. The thing that you should remember about FTD with the granulin mutation is this is one of these genetic diseases where a simple replacement of protein should be effective, because we know the proximal driving cause of neurodegeneration in patients with FTD-GRN is having too low of progranulin levels. Because these patients are haploinsufficient, they don't make enough progranulin. Low levels of progranulin leads to lysosomal dysfunction, a pathogenic inflammatory response, TDP-43 pathology, and ultimately, neurodegeneration. We know the proximal driving cause, which is low levels of progranulin. That is why we, with our program and others in the space, are seeking to replace progranulin. We know this hypothesis will work. It's who will have the best way of attacking it. What is the best mechanism that delivers the highest levels, that last the longest, to have the most effectiveness? We are in the midst of an ongoing Phase I/II study of PBFT02 in FTD GRN, primarily, but we've also opened it to FTD C9 patients as well. Let me focus on the FTD GRN population first, because that is our lead indication, and we will evaluate these two indications separately. The time course of FTD GRN is much more accelerated versus FTD C9. The evaluation of efficacy really needs to be separate for these two indications. We've treated nine patients to date who have FTD GRN across two dose levels. We've treated 7 patients as of this data release, with Dose 1, which is a higher dose, and we've treated 2 patients so far at Dose 2, which is half of Dose 1. It's actually a lower dose. We've moved to a lower dose for a couple of reasons: 1, because we've seen such excellent progranulin responsiveness, and 2, to maximize the risk-benefit of the product. We are currently enrolling patients in Cohort 3. Cohort 3 will be 10 patients, and they will all be treated with Dose 2. The key things we are looking at in terms of results are, of course, safety and tolerability, which we will discuss. We're looking for target engagement. First, these patients are missing progranulin. Their progranulin levels are too low in their CSF. How high can this product drive CSF progranulin? There's probably 3 different outcomes where we have natural history comparators, where we can look and say: "How is this product doing compared to natural history?" The first would be plasma neurofilaments, that is something that we have already shared data on, plasma neurofilaments. The second would be volumetric MRI, we've not shared any volumetric MRI data yet. The third would be the Clinical Dementia Rating scale. This is the ultimate primary endpoint of any registrational study. This measures clinical change, and it's a pretty large scale that has a measure of... it's a scale that's administered by physicians, and it's answered by patients and their families. Because of that, there's variability in the scale. Therefore, we need enough of an N and enough of a follow-up to really be able to see true differences versus natural history. What I'm going to share with you today is our preliminary data on progranulin and also plasma neurofilaments in comparison to natural history. This product is an AAV1 that is administered via the intra-cisterna magna route of administration. The cisterna magna is a large pocket of CSF that is located at the base of the skull. Direct delivery into the CSF allows for broad CNS biodistribution. We have excellent non-human primate data. I don't have time to share it all with you today, but it shows a distribution. It's in our corporate presentation. It shows a broad distribution in non-human primates who are administered this product via ICM administration. The vector is found throughout every region of the brain that has been sampled. It also allows us lower doses compared to IV systemic delivery and a decreased impact of neutralizing antibodies. Most importantly, this procedure takes less than an hour. It's done not by a neurosurgeon. It can be done by a neurosurgeon, but most of our procedures are done by interventional radiologists because it's a simple injection done under CT guidance. A bit about the baseline characteristics. This is from the first 8 patients who we've treated. The key thing I'm gonna call out here on baseline characteristics is, first, CSF progranulin levels at baseline range from 1.5-2.9. The normal level is 3-8. I will call that out on another chart as well. All these patients are a little bit below normal, but even that small decrease below normal drives all the neurodegeneration that they see. The other thing I will call out is the Global Clinical Dementia Rating scale score. The global score goes from 0 to 3, with 0 meaning completely unaffected and 3 being the most severe. 2s are moderately impaired, and 1s have mild impairment. For our first patients who we treated, we included CDR patients who had a CDR score at baseline of 1 or 2. Moving forward, Cohort 3 and beyond, we have changed our inclusion criteria. We are no longer including the more severe patients who have a CDR of 2, so CDR 2s are now excluded. We have opened up to even earlier symptomatic patients who have a score of 0.5. 0.5 is akin to mild cognitive impairment, where these patients are starting to show symptoms, but they haven't showed quite enough to be, have a full diagnosis of FTD yet. We know these patients are on the path to accelerate rapidly into full FTD once they show 0.5. Moving forward, we are treating an earlier population, 0.5s and 1s, because in neurodegeneration, our therapy and all the other therapies, we are not replacing neurons that have died. We are preserving neurons that are still functioning. It behooves us, if we want to see the most clinical efficacy, to get in to treat patients as early as possible. What I'm sharing here is some preliminary, very encouraging target engagement from our study. The y-axis you see is our target engagement, CSF progranulin. X-axis is time. This gray-shaded region is what I mentioned before, the healthy adult level of CSF progranulin. That is about 3-8, and you can see patients start slightly below the normal range. The top green line is dose 1. This is our higher dose. At this dose, we see a robust increase that peaks at about 12 months. Between 6 and 12 months, we know it peaks, and the levels are in the mid-20s, so markedly higher than the normal range. Our first data that we've shared so far, this is data we last shared in the midst of 2023, in June of 2025. Our first data point from our lower dose, which is half the dose of Dose 1, has already shown a robust increase to the high range of normal. We will be looking to the future for our future data release. The important thing that we're gonna be looking for is Dose 2. Our goal for Dose 2 is to have it settle somewhere in this white area above the normal range. People ask us, "Well, how high does progranulin need to get to see a clinical effect?" The answer is, nobody actually knows. It hasn't been shown anywhere because these are the first disease-modifying therapies that we've seen. Our goal as drug developers is to get as high of a level of target engagement as possible without seeing adverse events. There is no known toxicity from high levels of progranulin, and we have gone through all of our data with regulatory agencies, and they have no issue with us moving forward, treating patients even who start at a normal progranulin level, such as FTD C9 patients. I can say that other competitive programs have shown progranulin levels reaching about 4-5. There was an antibody approach that got to levels of about 4-5. There's also, I'll share some other competitive data on an AAV9 approach that has since been stopped, that has also delivered ICM, that showed an initial increase at month 2 into the teens, and then a decline by month 12 into the normal range. Competitively, we feel very strong that this is an excellent product profile for a progranulin replacement one-time gene therapy. I mentioned plasma neurofilaments as another biomarker that we are looking at. Plasma neurofilaments is one of the only disease progression biomarkers with longitudinal natural history data available. In terms of lab biomarkers, it is the only one. If you include volumetric MRI as a biomarker, there would be a couple. In two looks at the natural history data, on the left is an external study from the literature, and in the middle is our own analysis of existing natural history data. These are symptomatic FTD-GRN patients who are untreated. In untreated patients, their plasma neurofilaments increase at about a rate of 28%-29% a year. Whereas our four patients who have reached one year of follow-up have, on average, about a 4% increase in plasma neurofilaments. What we are seeing is that the product is showing a sign of slowing the rate of neurodegeneration in these symptomatic FTD-GRN patients. In terms of safety profile, PBFT02 was generally well-tolerated. There were 2 patients who had 3 SAEs, which were all asymptomatic and responded to treatment, all occurred at the higher dose 1. There were 2 episodes of venous sinus thrombus and 1 episode of LFT increase. The LFT increases, we have addressed by adjusting our immunosuppression regimen. Previously, this LFT increase was the first patient treated. That patient had only received oral steroids. We now give 3 days of IV steroids, followed by a 60-day course of oral steroids, and we've seen no more SAEs of LFT increase. This issue of venous sinus thrombus, a venous sinus thrombus is like a DVT, except not in your leg. It happens in one of the veins in these participants' brains. We attributed these as likely due to the product. Importantly, these patients had no evidence of thrombotic microangiopathy, or any other coagulation abnormalities, or any other signs of overwhelming inflammation. Our thought is these are likely local, a local inflammatory response, potentially from the capsid. The way we are attacking it is not to change the immunosuppression regimen. Moving forward in Cohort Three, all patients will receive a 5-to-6-week course of prophylactic anticoagulation. This is apixaban or Eliquis at half dose. This product, Eliquis, is approved for this indication, for DVT prophylaxis in the setting of older patients who are getting hip and knee replacement surgery. It has been well studied and has been quite effective in preventing DVTs. Competitively, this is where our product profile stands. At, we're a 1-time therapy. Given ICM, we get to levels of 26 nanograms per ml of CSF progranulin, and we've shown durability out to 18 months. This AAV9 gene therapy program was recently canceled. The program was stopped. One of the big issues with this program was a declining CSF progranulin level from month 2 to month 12, as I addressed before. There was another AAV9 gene therapy program that is also delivering a gene that generates progranulin. This is delivered via the intrathalamic route. It requires neurosurgery. It is done with a device that we have studied also for our Huntington's program, and this device requires quite a long neurosurgery, basically all day in the OR with a neurosurgeon. We like where our delivery and our target engagement stands versus the competitive space. One note: we are also treating or using PBFT02 to treat patients with FTD-C9orf72, in this case, the rationale here is not replacing low levels of progranulin. The rationale here is that there is a final common pathway of many different diseases. FTD-GRN, FTD C9, also sporadic ALS all have TDP-43 pathology as a final common pathway. TDP-43 is a protein that should be in the nucleus. With this pathology, TDP-43 mislocalizes from the nucleus to the cytoplasm and forms inclusion bodies that are associated with neurodegeneration. We have in our corporate deck the evidence. There is ample preclinical evidence that raising progranulin levels to levels higher than the normal range has been shown to ameliorate TDP-43 pathology in preclinical models. Thus, we are treating patients with FTD-C9 in our ongoing trial. These are two separate cohorts, Cohort Four and Cohort Five, and Cohort Four has started already. One note on the Huntington's program. We have recently shared what our target is for the Huntington's program. Huntington's, as I'm sure you all know, is caused by CAG repeats. In Huntington's disease, the length of the repeat can elongate over time. It's a term that is called somatic instability. CAG expansion above a certain threshold leads to neurodegeneration. Longer repeats are associated with worse disease pathology. CAG expansion occurs at different rates in different neurons and is fastest in the caudate and putamen brain regions, which degenerate first in Huntington's disease. DNA repair proteins play a key role in CAG repeat expansion, and MSH3 is a DNA repair protein that is a key driver of somatic instability. In the presence of certain CAG motifs, MSH3 can make a mistake. It can erroneously cause CAGs to be incorporated instead of snipped out, leading to an expansion of the repeat sequence. In patients with Huntington's disease, those who have certain variants in MSH3 that alter MSH3 function are associated with delayed onset and slower progression. In Huntington's mice, MSH3 is essential for CAG expansion, and knockdown has reduced somatic instability and reduced Huntington's pathology. MSH3 is an exciting new target, and this is what we are attacking Huntington's with. We are going to be delivering via an AAV, a miRNA that will knock down MSH3. We plan to utilize an optimized intraparenchymal delivery approach. In Huntington's disease, it's different than FTD-GRN. We actually need to get the vector into neurons that are at risk, so we have to use an intraparenchymal approach. We want this one-time intraparenchymal approach to not take up an entire day of a hospital's operating room time. Our procedure will have reduced procedure time. We have completed our proof-of-concept studies. We're really happy with the amount of knockdown that we get and the amount of somatic instability that we have reduced, but we have additional preclinical studies ongoing, and we expect to declare a clinical candidate in the second half of this year. Looking ahead, this is my final slide, and then I'm happy to answer any questions. The first half of this year is a big moment for us, for the FTD-GRN program. We will be reporting updated interim safety and biomarker data, not only from Dose 2, but also data from Dose 1 as well. The Dose 2 data, as I mentioned, is important. We will be sharing our 1st 6-month data from Dose 2 in the 1st half of this year, but we'll also be sharing more longitudinal data on the Dose 1 patients because they were the 1st patients treated. We're seeking regulatory feedback on registrational trial design in FTD-GRN. There's many options we have for registrational design. There's been uncertainty recently. We all have seen it with the agency, what we want to understand now, before we actually do the registrational setup, is with this FDA, what are the requirements to get to approval? Any outcome is fine with us. We plan for a variety of outcomes. The important thing is for us to know now what it will take to get the product across the finish line, so we can plan for that. As I mentioned, we'll declare a clinical candidate in the second half of this year. We have a cash balance as of the end of the year of $46 million. We have a current burn of between $30 million-$35 million, so we have runway out into the first quarter of 2027. Clearly, these data inflections in the first half of this year will be important to the company. With that, Jay Olson, I'm happy to pause there and answer any questions as needed. Great. Thank you so much, Will. Really appreciate your bringing us up to speed on all the impressive progress you're making at Passage Bio. Yeah, we do have a few questions. Okay. I'm glad you mentioned some of the competitor programmes in FTD. Maybe starting off with the recent for Alector. Can you just talk about any potential implications or read across to PBFT02, and any learnings from Alector's programme that you can apply to your own programme in FTD? Absolutely. Three things on the Alector program. With Alector, unfortunately, there's no data that has been shared, so we don't know. We only have hypotheses for why the trial might not have worked. The first is mechanistic. The Alector product was blocking sortilin. Progranulin, in its normal environment, needs to get from outside the cell into the cell to exert a positive effect inside the cell on the lysosome. The Alector product was blocking sortilin. Sortillin shuttles progranulin for destruction, but it also shuttles progranulin to the lysosome for a positive effect. Due to this mechanism, it's not clear, was the product also capping the positive effects? Was it limiting the amount of progranulin that was actually getting into the cell to get to the lysosome? We have no way of measuring it. All we are measuring, we're measuring extracellular progranulin through CSF. That's what everyone is measuring. No one is measuring what's in the cell. Perhaps the mechanism was not the appropriate one to attack this disease. Hypothesis one. These are only hypotheses. Hypothesis two, I mentioned before, is perhaps the progranulin level was not high enough. From the phase 2 data that we saw from the Alector product, this product got progranulin levels in the 4 to 5 range, perhaps that wasn't high enough to affect enough patients. We don't know. The final question is, perhaps this was not the right population treated. As I mentioned, you wanna treat patients as early as possible. We have, moving forward, limited all of our treatment to only patients who are CDR 0.5 and 1. We have now excluded twos because we don't think they are likely to benefit as much from our product. The Alector study, 30% of those patients were CDR twos. That study was capped. There was a limit of CDR twos because all of us studying neurodegeneration know that the risk of not seeing a clinical effect is higher in patients who are more progressed. We are not including any of those patients, so our percentage of CDR twos moving forward will be 0. Okay, that's super helpful. Mm-hmm. Thank you for explaining that. I wanted to follow up on your comment about some additional safety data, biomarker data-. Mm ... that you'll be disclosing in the first half of the year for Cohort 2 in FTD. Can you just talk about what investors should be looking for there, and maybe kind of... Yeah, yeah. ... frame expectations? Thank you. Sure. A couple things. First for dose one, we will have more 12 and 18-month follow-ups, so looking for continued durability, so more of the same is good. Dose two is very important. The first six-month data point, you can see the pattern. We expect the product to keep rising. If it peters out in this healthy adult range, that's gonna look more like the AAV9 Lilly program in terms of low levels. That wouldn't be as good. We hope for it to be somewhere in this white range. It's important because the closer the dose two progranulin levels get to dose one, because our dose one patients have the longest follow-up, they're gonna have the most clinical biomarker. For instance, their neurofilament data has... We have the most neurofilament data on them, they're gonna have the most longitudinal follow-up. The closer that dose two gets to dose one, the more in the future we can bridge the clinical data between dose one and dose two. The closer dose two gets to dose one, it allows us to make the decision of what the dose will be easier. If they're very close, then it makes it very easy, if you can have similar target engagement with less potential for side effects, because clearly the side effects from an AAV are correlated with the dose that you give. Okay. That would be an important thing. The other thing I would look for is a refresh of the neurofilament data. With a larger N at the 12 months, are we still having an effect on neurodegeneration? To the extent that there are any other biomarkers that we can show a change from natural history, we would consider sharing that as well. I would not expect CDR data. If you look at particularly even all of our competitors who've ultimately shared CDR data, they had a data set of at least 12 patients or so, followed for about 1 year, and we are certainly not at that point yet in our follow-up. Okay. Makes sense. That's helpful. Thank you, Will. We'll look forward to that. Then, maybe, one last question. Mm-hmm. I'm glad you mentioned your Huntington's program, and sounds like you're making a lot of progress there. Could you just talk about the clinical profile that you hope to achieve and some of the differentiating features versus other programs? Yeah. Yeah, sure. I mentioned one of the differentiating features, which is the time of administration and the methodology for administration. This is important, especially when you're thinking about commercial utilization and uptake. One thing about the DNA repair protein pathway, when you compare it to, say, knocking down mutant huntingtin, is, this is something the expansion of repeats is something that happens earlier. When you have this type of approach, attacking DNA repair proteins, there's the potential to actually treat patients earlier on, potentially even prodromal patients, earlier on in the disease course, so that you're actually preventing them from having these long CAG repeat sequences and making mutant huntingtin protein. When we talk to experts in the field, they see there's room for multiple approaches in Huntington's disease, particularly when you look at different phases of the disease. Okay, we'll look forward to your clinical candidate nomination in the second half of the year. What's the timeline for moving that into the clinic? Yeah, Jay, I'd love to share that, but we haven't given that guidance so far. The first guidance is gonna be clinical candidate selection, and what we're really looking forward to doing is sharing the suite of preclinical data that we have that's getting us excited about this program. Okay, excellent. All right, we'll wrap things up there. Thank you so much, Will. Really appreciate your making time to share all the impressive progress with us that you're making on behalf of these patients with such high unmet medical needs. Really, really great to see, and we'll look forward to future updates. Great. Thanks for having me, Jay. Take care. Our pleasure. Thanks, everyone.