All right. Okay. Welcome back to one of the last sessions of the TD Cowen's 46th Annual Healthcare Conference. I'm Brendan Smith, one of the Biotechnology analysts here at TD, it is my pleasure to introduce CEO of Passage Bio, Will Chu. He's gonna give a brief presentation up front, then, excuse me, as is the case for most of our fireside chats and presentation slots throughout the conference, if you've got any questions, maybe kind of hold off towards the latter end, or if you wanna send them to me at brendan.smith@tdsecurities.com kind of work them into Q&A at the back end. For now, I will hand it off to Will.
Okay. Thank you, Brendan. Excited to close things out for us at the conference. William Chou, CEO of Passage Bio. I'm gonna be talking today primarily about the first thing listed here, which is our clinical stage AAV program in FTD with a granulin mutation that we believe is quite differentiated in this market. We also have a preclinical program in Huntington's disease that I'll go through our target approach there, and we have cash runway through the Q1 of 2027, so plenty of opportunity to keep turning over more data cards. Okay. The thrust of today's discussion will be on FTD-GRN. This is our clinical program. Frontotemporal dementia with the granulin mutation affects about 18,000 patients across the U.S. and Europe. This is a rare disease, but it is certainly not an ultra-rare indication. What is FTD?
It is a early-onset behavioral dementia. These patients not only have memory loss, but they have a variety of behavioral abnormalities that is untreatable right now. There's no interventions at all, and patients generally live on average about eight years from diagnosis, so it's rapidly progressing. Now, in the specific genetic subtype of FTD patient who this program is treating, they have a mutation in the granulin gene. Due to haploinsufficiency in the granulin gene, they end up having very low levels of the protein progranulin. Low levels of progranulin is the driving cause of neurodegeneration in these patients. If you don't have enough progranulin is critical for lysosomal function, so you get lysosomal dysfunction, pathogenic inflammation, TDP-43 pathology, and ultimately neurodegeneration. This is a relatively straightforward genetic medicine approach. These patients are lacking enough protein progranulin.
This product, this AAV product, gives them replacement for that to generate more progranulin. A little bit about the homeostasis of progranulin and how it works in the normal environment. Progranulin is excreted into the extracellular space. It is then, in its natural environment, endocytosed by a variety of cell membrane receptors into the neuron. It needs to get inside the neuron to exert its positive effect on lysosomal health. Our approach and others' approach is to drive greater production of progranulin such that we have more substrate available for these receptors to endocytose. This is a schematic of our ongoing phase 1/2 study in FTD-GRN. We have treated 9 FTD-GRN patients to date and 1 FTD-C9 patient as well. We've completed cohorts one and two. We have enrolled the first three patients in our final cohort of FTD-GRN patients, which is cohort 3.
The important change that we've made a couple of changes between cohorts two and three, but one important change that I'll cover here is that we've modified the inclusion criteria. All our patients treated so far in cohorts one and two, they were split between both mild and moderate FTD-GRN patients. Moving forward, we have now narrowed the inclusion criteria, so we are excluding the more moderate patients, and we are including a broader population of mild and mild cognitive impairment patients. We're basically moving the target population to an earlier target population moving forward. In neurodegenerative disease, none of our therapy and other therapies are not replacing lost neurons. You need to catch patients early enough to really have a potential to have impact on clinical outcomes. Some of the endpoints that we're looking at, of course, safety and tolerability.
Progranulin, particularly CSF progranulin, that's our measure of target engagement, so we do wanna see how high this product can get CSF progranulin. A couple of other secondary endpoints to call out. Plasma neurofilaments and volumetric MRI are really the only two biomarkers where there's good natural history data where we can compare treated patients to the natural progression we would see in untreated symptomatic FTD-GRN patients. To date, we have shared some early data around plasma neurofilaments, and I'll show that to you in a couple of slides. Ultimately, we do wanna measure clinical change, and that's measured by the Clinical Dementia Rating scale, sum of boxes. One other thing to note here is that we are testing two different dose levels. The first dose we tested is actually our higher dose.
That's 4.5 to the 13th total genome copies administered. Overall, compared to other gene therapies, this is a relatively low dose. Now we're also testing an even lower dose, so half of the first dose. All the patients in cohort three will be treated with dose two, which is our lower dose. I'll share with you for both efficacy and safety reasons why we've moved to a lower dose. How is the product administered? Again, this is an AAV1 product. It's a 1-time therapy. We administer it through the intracisterna magna route. The cisterna magna is a pocket of CSF that is at the base of the skull but is outside the brain. From an anatomic perspective, this is similar to the intrathecal route. It's just closer to the brain.
We get more vector into the brain this way than intrathecally. Importantly, this is a brief administration. Total time is less than 1 hour. It's done either by a neurosurgeon or in most of our cases, by an interventional radiologist. It's done under CT guidance. Unlike other procedures that we've heard about, which are quite long or in the operating room for 10-12 hours done by a neurosurgeon, this is not. This is one hour by an interventional radiologist. Some initial data from our study. What we're showing here is target engagement. The Y-axis is CSF levels of progranulin. That gray shaded area is the normal level of progranulin in unaffected patients. Normal is about three-eight. These patients are haploinsufficient. They start at a level of about 2 to 2.5.
You can see those dots start below the gray area. The green is our higher dose patients. We've treated more patients with a higher dose. You can see that the higher dose of this product gets very robust elevations of progranulin. The levels peak somewhere between month six and month 12, and they've consistently been up to the low 20 range. We've only shared one data point as of this data cut, which was mid-2025, of the lower dose, and that was very early. It was only at day 30, but already at day 30, that patient had gotten to a level at the upper end of the normal range. Based on the pattern that we've seen from the higher dose, we expect that to go up over time.
We will have new data coming out in the H1 of this year, some important data for us is more longitudinal data at this lower dose. We'll have some six-month data at this lower dose. The question for us is, our target is to get on average above the healthy normal range. Are we starting to trend in the right direction there, or does the six-month level look low towards that healthy adult range? We're asked all the time, how high do you need to get progranulin to actually have a clinical effect? The answer is, nobody knows. When nobody knows, and it could be different for every patient, our goal is to get the target engagement as high as possible without seeing adverse events.
We have had 3 SAEs at the higher dose, that's one of the reasons why we're moving to a lower dose. I will also say that there's other therapies. There is an anti-sortilin antibody therapy that was studied in a very large randomized controlled study. That product only elevated progranulin into the four-five range based on phase two data. It appears that at least that level of progranulin elevation does not seem to be enough. We definitely wanna get higher than the gray area. I mentioned plasma neurofilaments as one of the biomarkers where there's good natural history data. On this slide, you can see two different looks at the natural history data. On the far left is from the literature in symptomatic FTD-GRN patients, and the middle gray bar is from our own analysis of the natural history.
Both of them show pretty similar rates of change of plasma neurofilaments over one year. It increases by about 29% a year in symptomatic FTD-GRN patients. Our first four patients who've reached one year of follow-up had an increase of only 3.7%, so relatively flat. At least what we're seeing at this very early point in our phase one study is a slowing of neurodegeneration. In terms of the toxicity profile, I did mention at the higher dose, we had 3 SAEs in two patients. The SAE of most interest was the venous sinus thrombosis that 2 patients had. A VST is essentially a DVT, but it happens in one of the veins of the brain. These were asymptomatic and responded to treatment, which is anticoagulation.
They were just picked up on MRI. Because we had these SAEs and because they were likely attributable to the product, we've done two things. We've moved to a lower dose, and we've not seen any SAEs since moving to a lower dose. We've treated now three patients at the lower dose. We've also introduced for cohort three, a six-week course of low-dose prophylactic anticoagulation. This is Eliquis at half the therapeutic dose. Eliquis is already approved for this indication, prophylaxis of DVTs in patients who are undergoing hip and knee replacement surgery. It's been very widely studied in an older population, so we felt very comfortable giving this regimen as prophylaxis. A bit about the competitive space. PBFT02, I never mentioned the name of the asset, but PBFT02 is our AAV1, and you can see the profile right there.
It's a one-time therapy. It's gotten to levels of 26 at our higher dose at 12 months, and it's shown great durability out to 18 months. There's also an AAV9 gene therapy that is administered directly intrathalamically. That is a relatively long procedure. Other programs that have used the ClearPoint technology to do intraparenchymal administration have talked about operating room times in the 10 to 12 hours. That requires a neurosurgeon, whereas ours is a less than one-hour procedure that requires an interventional radiologist. There's also a progranulin replacement therapy that is in phase one, two that it requires chronic dosing. If it's every month, that means the patient and their family with a behavioral dementia will be going to the hospital or clinic once a month forever.
That is our clinical program. Just briefly about our preclinical program. I think you all know how Huntington's works. The problem with Huntington's is patients have CAG repeat tracts that are too long, greater than 35, really leads to the formation of a mutant huntingtin protein. What we found is that the extension of that repeat over time is what causes a fair amount of the morbidity. CAG expansion leads to neurodegeneration. Longer repeats are associated with worse pathology. The CAG expansion occurs at different rates at different parts of the brain, and where it happens the fastest is where all the degeneration starts, in the caudate and putamen. There's been a lot of research in the last couple of years on DNA repair proteins.
MSH3 is a DNA repair protein that should be repairing DNA, but in the presence of certain CAG repeats, it does the wrong thing. It actually incorporates more CAG repeats into the DNA, so more MSH3 is bad. We have seen in Huntington's patients who have a genetic variant that decreases MSH3 function actually do better. They have delayed onset and slower progression. Preclinical models have shown that MSH3 is really important for CAG expansion, and knocking it down can slow down somatic instability and slow Huntington's pathology. It's an exciting new target. We are going to be knocking down MSH3 via a microRNA approach, a vectorized microRNA. It's gonna be a one-time delivery. Importantly, our approach will not require a prolonged procedure.
We're using a different procedure with a markedly reduced procedure time because we think that is going to be important ultimately for commercial uptake. We finished our proof of concept studies. We're excited by what we're seeing in terms of reduction of somatic instability, and we'll be declaring a clinical candidate in the 2nd half of this year. The last slide before I do some Q&A is what to expect. We've guided to refresh our data the 1st half of this year. We're excited to do that. We also have guided that we will be seeking early regulatory feedback on the registrational path for FTD-GRN. 2nd half of this year, as I mentioned, we'll declare clinical candidate selection for Huntington's disease. I'm sure there's plenty of questions on regulatory. I know Brendan has them.
There's a lot of news these days about regulatory feedback. Happy to move into the Q&A section. All right.
All right. All right, great. Maybe before we actually dive into Huntington's, of which there will be, I'm sure, no shortage of things to discuss, maybe help us understand, you know, you referenced a few of the other approaches within PGRN that have had some data readouts.
Mm-hmm.
Some hiccups, using other mechanisms.
Sure.
maybe help us kind of extrapolate, like, what is it in your view that kind of led to the results, whether that's the Alector asset or the Prevail asset or any other approaches that really kind of set apart what Passage Bio is doing in this space?
The Elekta asset was an anti-sortilin antibody that did not meet its endpoints or show any improvements in any biomarkers in a phase three RCT. We don't know 'cause we haven't seen the data, right? We only have hypotheses for why that product might not have performed. The first is mechanistic. Blocking sortilin, you're not actually creating any new progranulin. You also are not guaranteeing that progranulin is getting inside the cell where it needs to get. Mechanistically, there could be an issue versus just plain old replacement adding more progranulin, which is what other approaches are doing, us and others. There's a mechanistic question. There is a level of progranulin issue. I think I mentioned that when I covered our data.
Perhaps getting to a level of four to five, just normalizing on average, is not enough for patients who've already started with a neurodegenerative process. The last could be patient selection. In that study, about 1/3 of the patients were the more moderately impaired patient. As I mentioned, we are now excluding all moderately impaired patient and only including earlier patients who are earlier on in their neurodegenerative process. If you don't select the right patients, if you include patients who are too late to really help, you're not gonna see a positive study. I don't know which of those it was, but it could be one, it could be some of all of the above.
I mean, in your last point there, just about actual patient selection, does, the patients that you're now focusing on.
Mm-hmm.
moving forward, does that impact anything to do with the actual, duration of clinical endpoints that we need to be looking at? Like, if they're earlier on in disease course, do you need to be monitoring them longer? Have you.
No. In fact, if you look at the Staffaroni paper, that's the most famous paper that has used all the natural history data. There is a very rapid acceleration. As soon as patients start showing symptoms, they rapidly decline. As they progress more in their course, things start flattening out. No, these patients, everyone who is coming into our study is already showing symptoms. If we were studying asymptomatic patients, then you might have to wait a long time. We don't have any markers for knowing when people are actually gonna phenoconvert. That's why it's difficult to start in completely asymptomatic patients.
Okay. Just to kind of put a pin in the mechanistic differentiation, I guess. Do we have any sense of the impact of using, you know, AAV1, for example, like you guys, versus an AAV9 or anything like that? Do we know kind of mechanistically what those are targeting within the brain and how that could impact?
Right. You mentioned the Lilly program. Lilly just canceled their program, which was an AAV9 that also was progranulin replacement that was delivered directly just like ours, ICM. A couple things. In our non-human primate studies, AAV1 did dramatically better than AAV9 in generating progranulin. Some of that we think may have to do with increased tropism that AAV1 has for certain cells called ependymal cells. They are forever cells that line the ventricles and generate CSF. That could explain why AAV1 in general creates more progranulin. There was another feature of the Lilly Prevail product that I don't think had anything to do with AAV1 or AAV9. The progranulin profile declined over time from month 2 to month 12.
We've used AAV9s in some of our other programs, in our legacy pediatric programs, and we never saw a decline like that. Others have studied AAV9s in CNS and have not seen declines in target engagement like that. There's something beyond just the capsid that was happening there.
Interesting. All right. Just to kind of level set the expectations for the data update.
Right.
I had, it sounds kind of like it's, you know, longer duration follow-up. Wanna make sure that the safety protocol is still kind of keeping up with what you needed to be doing. In the context of this year, is kind of no news, good news type of situation? Like, if it's kind of consistent with what you saw at the earlier cuts, equally as safe and durability of the actual impact is looking comparable to the prior cut. Is that kind of checking the box there for you guys?
Yeah, I think that's very important. No news is good news from a safety standpoint. Consistency, as you mentioned. To the extent that there are other signals, we would love to share those as well of potential efficacy. The other thing that I would also look at is we have narrowed the target population to earlier patients. Are we seeing any difference between earlier and later patients even at this early stage?
Okay. Got it. All right. I did wanna ask a little bit about the conversations with FDA-
Sure.
that are ongoing, right? I know when we caught up recently, I think that's gonna be an important update also.
Right.
What are you kind of hoping to will come out of these conversations with FDA, and what should we kind of be watching for as, like, this is kind of ultimate arbiter of success in terms of what you're trying to get out of that?
Right. I think we just want certainty. I think certainty has been a little hard to come by, and we want certainty now, which is very early. We have not started our registrational study. We are in the planning stage now. We wanna know what the bar is, what we can actually plan for. We will approach with two ideas. We could do a single arm study. We could do a randomized controlled study. Both are possible with this intervention in this space. We will make a case for a single arm study because the FDA has given some guidance that way. I would say there's been a fair amount of recent news from the agency, whether it's with REGENXBIO or with uniQure, that they've had issues.
They've had issues with the design of studies, with comparisons to natural history. It's hard to know because there are some words, some things that the FDA has said, there's actions that they've actually done as well. We won't be surprised by any response. We're prepared to discuss a couple types of trial designs, and to talk with them about how best to design either way.
Just to confirm, because I know you've also started, enrolling and dosing in the C9orf72 cohort.
Yes
...with the dose two.
Right.
As of now, is kind of the plan, the go forward plan with FDA for registrational study to focus just first on the FTD-GRN population? Or would you potentially have a separate cohort of C9orf?
Right now we're gonna focus on FTD-GRN for those discussions with the FDA. When you evaluate efficacy, you have to evaluate GRN and C nine very differently because the pace of the disease is slower in C nine versus GRN. However, the safety profile of the product is the safety profile of the product. The patients who we treat with FTD-C9, their safety profile is gonna be very important for that discussion with the agency on the safety of dosing patients at any given dose.
All right. From this phase one, two for cohorts three and four, do we have a sense of like, where are we kind of in the enrollment process of this and, is it fair to kind of expect at least preliminary data from these maybe later this year or next year?
We've enrolled the first three patients in the FTD-GRN cohort three. We've treated the first patient in FTD-C9. The enrollment has been brisk so far. Three in the Q1 is great. We've seen a lot of interest in progranulin replacement from patients since all the news that we talked about with other studies. That's very positive in terms of momentum of enrollment.
And I guess that is kind of a underappreciated aspect to this, right? 'Cause I think for a while, you know, maybe six, seven years ago, it felt like there were a lot of new approaches moving into FTD-GRN.
Mm-hmm.
I think because of that, there felt like there was a little bit of a slowdown in everybody being able to enroll what is, you know, an orphan indication.
Sure.
As, you know, to your point, as some of these other studies have been discontinued or gone by the wayside, are you noticing a pretty material uptick in either location willingness, or enrollment rate within the population?
oticed that. We've seen it in our data, and we've also heard it anecdotally from our investigators. There is a natural interest for patients to maybe wait. If something is about to get approved, then there's going to be an approved product versus joining a study. Maybe they're waiting for an approved product. As soon as that study failed, our investigators saw an increase in interest in our study. And so I would say the same from a site standpoint, after the Lilly study decided to stop.
Got it. All right. Maybe just in the last couple minutes, I wanted to talk a little bit more about, the HD program.
Sure
...but also how you're kind of thinking about the development path there, given kind of the updates related to uniQure and what the messaging out of FDA is. You know, irrespective of the mechanism, I guess, how is the development path for Huntington's for you guys distinct from what you've kinda gone through with FTD?
Yeah. I'll answer this in a couple of ways. I would say the where we are right now with somatic instability, the status of early biomarkers, it's earlier on than, say, all of the data we have in FTD-GRN. It would also be a new therapeutic area for us. It may not behoove us to go on our own in this area, in Huntington's disease. In particular, we know there is a lot of interest in the DNA repair mechanism overall. That's why even though we're not sharing data yet, we are sharing what our target is because we know that it's important that this is an area of interest for others.
Is there plans at some point before you move into a phase I to potentially share some of the preclinical data?
Definitely
maybe next year? Yeah.
Definitely.
Okay.
Yeah.
All right. As I, as we kind of look at the, you know, you mentioned potentially not going it alone. What would kind of a ideal, whether development or commercialization partner look like? Admitted that it's early, but, how does that kind of factor into your thinking about how to progress these two programs?
Yeah. I mean, ideally, we have two programs in fairly large neurodegenerative disease indications. These are gene therapies, so you need to have a commitment to this space, right? That is the number one thing, because we are one company, we have two main programs, and they're both in the same space. I think that is the most important piece. I've heard along the way, right, some companies like gene therapies, some are less into gene therapies. The way I look at, particularly our FTD GRN program, the gene therapy, the part that is different from every other type of modality is really the manufacturing. We haven't talked too much about manufacturing, but we've covered the manufacturing.
We're able to make a very, we have a very efficient process to make this at scale, and we've had numerous discussions with the FDA about it, and we have a great path forward to manufacturing. We don't manufacture it ourselves. It can be manufactured at our, with our partner at Catalent. Really it's just about clinical data then and doing a study. You take the whole manufacturing part of the gene therapy off the table. I think the biggest thing for us is a commitment to neurodegenerative disease.
Got it. That manufacturing process equally applies to both your programs, or is it just FTD for now?
It's just FTD right now.
Yeah. Okay.
Yeah.
Got it. All right. I think with that, we are right at time. Wanted to thank everybody for listening in, joining us all the way to the end of the conference. Really appreciate it. Will, it's always good to see you. Thanks for joining.
All right. Thanks, Brendan.