Maze Therapeutics, Inc. (MAZE) September 9, 2026 Conference Transcript

All right. Hi, everyone. Thanks for joining us at the Wells Fargo Health Care Conference. My name is Sadia Rahman. I'm one of the biotech analysts here at Wells Fargo, and it's my pleasure to introduce Maze Therapeutics for our next session. From Maze, we have Misbah Tahrir, the CFO. Thanks, Misbah, for joining us.

Sadia, thanks for having us at the conference. It's great and exciting to represent Maze this morning.

Yeah. So I'll hand it over to you for any opening remarks, and then we'll get into Q&A.

Well, let me start with a little bit of background on Maze for folks who might be newer to the story. Maze is a clinical stage biopharmaceutical company focusing on the development of precision medicines for patients who suffer from kidney and metabolic diseases. We have three programs in clinical development. The first is MZE-829. It's an APOL1 inhibitor for the treatment of patients who suffer from APOL1-mediated kidney disease, or AMKD. We're currently testing 829 in a phase 2 study called Horizon. During the course of the fireside chat, we'll be talking a lot about the Horizon study. Our second acid is MZE782. That's an inhibitor of SLC6A19. That's a transporter of neutral amino acids, and we're testing that. We just announced that we initiated a phase two study to test that in patients who suffer from PKU. Our third program is also MZE782, but we're also testing that drug in patients who suffer from chronic kidney disease, or CKD, and we've guided to initiating that study in the first half, that phase two study, in the first half of next year. All of those drugs are small molecules. They're wholly owned by the company, and they're driven by the support of our Compass platform, which is our drug discovery engine, in which we harness the power of human genetics to develop precision medicines. Great.

That's a great overview. So you had your Horizon readout in the first quarter. Going into that readout, a lot of the conviction for this mechanism came from human genetics data showing it, you know, suggesting it can work in broad AMKD, suggesting that APL1 drives faster progression across different kidney disease phenotypes, including in patients with diabetes. So from the first readout of Horizon, has your view changed of where APL-L1 inhibition can work or might be more relevant across the different etiologies? You know, where are you more confident coming out of that readout? Maybe where are you less confident?

Maybe a little bit of background on Horizon and APL-L1. If you suffer from AMKD, it starts with the genetics. If you have two copies of the APO1 high-risk variants, you're at risk for developing a more aggressive, more accelerated form of chronic kidney disease. You may develop CKD at an earlier age, likely perhaps before you turn 50, and you may progress much faster and potentially even get to dialysis 10 years earlier. So we're testing eight to nine in a phase two open label basket study of proteinuric CKD adults who have two copies of the APOLO1 high risk variants. And we're testing across a spectrum of AMKD patients. So AMKD patients who have FSGS, as well as patients, AMKD patients with diabetes and AMKD patients without diabetes. And certainly we're testing for safety and tolerability. Most are also looking to see what level of pertinoria reduction we can achieve with MZE-829. Back in March of 2026, we released some initial data from the Horizon study. We had 12 evaluable patients, and the results were positive. Not only was 829 well-tolerated, but we saw an average UACR reduction of 36% across the 12 evaluable patients. Half of those patients achieved a response, and we're defining response as a 30% or greater UACR reduction. And that's a threshold that's been established by KOLs based on research. There's linkages between that threshold and improved kidney outcomes, long-term benefits. So we set that as the bar that we would look at going forward. So we saw a 36% average UACR reduction in half the patients achieving a response. In FSGS patients specifically, there were four of them in that initial data set. We saw an average UACR reduction of 62% with three out of the four patients achieving a response. But perhaps most importantly, we saw responses across each of the three patient segments, AMKD patients with FSGS, AMKD patients with diabetes, and AMKD hypertensive patients without diabetes. We saw responses across each of those three segments. And so, really, the potential is still very much there for MZ-829 to work across the broad spectrum of AMKD patients.

You mentioned FSGS patients. They had particularly deep responses in that readout. And some investors wonder if FSGS patients might respond better to these drugs. Is there any biological reason to think that APL1 inhibition could work better in FSGS than in other forms of even non-diabetic AMKD? You know, could maybe AMKD damage be associated with FSGS histology?

I think it's too early to make conclusions based on that initial horizon data set. but we certainly liked what we saw in that data set across the range of patients. From a biology perspective, we know that if you have AMKD, you're more likely to develop a more aggressive form of CKD and to progress faster and reach dialysis faster, irrespective of where you are on that spectrum. So we see that faster progression in patients with diabetes, without diabetes, with FSGS, without FSGS. So APOL1 impacts all those different patient segments. We also know that there's a protective variant called N264K that confers protection for these patients irrespective of their diabetes status or FSGS status. And we've used that variant to inform the design of MZ829. So in theory, and based on what we saw in the initial data set from Horizon, we think A29 could still very much work across those segments of patients.

And on FSGS, you had previously estimated that FSGS comprises only around 5% of the addressable AMKD population, but these trials in AMKD seem to be enrolling more FSGS patients than expected. Could that mean FSGS is more common or underdiagnosed in AMKD than previously thought, or could this just be more of a feature of how the trials are enrolling?

So we talked about the three different segments that we're testing in Horizon, AMKD patients with FSGS, with diabetes, hypertensive without diabetes. We estimate that there are 250,000 of such patients in the United States. AMKD predominantly impacts people, starting with the genetics, going back to the genetics, impacts people of West African descent, 250,000 potential patients in the United States. There are no approved treatments today. FSGS, we've estimated to be roughly 10,000 patients or so, but it's possible that folks have underestimated that patient population. Recall that it is a biopsy-driven diagnosis, so it's relatively clean in that respect. But as we're finding as more and more patients get biopsied as awareness around FSGS with newly approved treatments and awareness around AMKD grows, it certainly wouldn't be surprising if we see that estimate of FSGS patients grow over time.

Great. So you have some preclinical data that suggests that MZ829 is more potent than the competitor compound from Vertex and Axoplin. How would you compare the efficacy observed in Horizon so far to what was reported in the Axoplin readout in FSGS patients, and do you see that thesis that, you know, your compound is more potent could lead to better efficacy playing out in the data?

Yeah, we'll certainly let Vertex speak to their drug, but the potential is certainly there for MZ829 to be a potent treatment for AMKD patients. We know from our preclinical work that MZ-829 may have a dual mechanism of action. So APL-L1 works, if you have the high-risk variants, you're more likely to develop a form of CKD where the APL-L1 is punching holes in the podocytes of the kidney. And 829, our drug, could work by not only blocking those holes, those pores, that form, but also disrupting the formation of those pores in the first place. So that could make for a pretty potent combination, pretty effective approach to treating AMKD. So in the clinic, in the initial horizon results, as I mentioned, we saw an average UACR reduction of 36% across the 12 available patients. And then with FSGS, we saw a 62% average UACR reduction. So that certainly suggests that we may have a potent drug in AMKD, but we clearly need to test more patients to confirm that effect. And that's what we're doing right now with Horizon. We're continuing to enroll patients, and we're looking forward to presenting more data end of this year, early next year.

So after the first readout, you lowered UACR entry threshold to 200 mg per gram and also capped FSGS enrollment. What led you to make those adjustments? And, you know, what are you specifically trying to learn from the next update with those changes?

Yeah, we definitely try to incorporate learnings from our clinical work to date. And in the initial data release from Horizon, we didn't necessarily see a simple relationship between baseline perineal levels and response or efficacy. So we felt comfortable tweaking the threshold for enrollment slightly from 300 to 200. Recall the normal cutoff for proteinuria is 30. So even above 30, you're getting to abnormal levels of proteinuria. So we lowered the threshold from 300 to 200. It may also help with enrollment. We'll see. We also capped the number of FSGS patients just to make sure we don't over-enroll in one segment to the detriment of the others. We're guiding to releasing data in 10 to 15 patients in each of the three segments that we've been talking about end of this year, early next year.

And I think you've said more enrollment is coming from EU sites. And with these updates to the inclusion criteria as well, could the population characteristics look different in in terms of disease stage, lower proteinuria, or different background therapy that could influence the outcome?

You know, it's certainly possible the data set we're gonna release end of this year, early next year, will have in total inclusive what we've already shown, you know, three to four times the number of patients we showed in the initial data released back in March. So there certainly could be differences, but there's nothing to suggest so far that our expectations should change or they should see dramatically different patient characteristics.

And then for the phase three design, you're targeting a trial start in the first half of 2027. Is the Horizon readout gating to that, you know, ahead of the phase three starting? What could you learn from Horizon or Vertex's readouts that, you know, could maybe change that decision to proceed only with non-diabetic AMKD? or is that the plan to go forward and only non-diabetic?

Yeah, so we came out of the March 2026 readout from the initial data from Horizon with two clear priorities. One was to complete the enrollment on Horizon and generate additional patient data, and that's what I referred to that we're going to release end of this year, early next year. The second priority was to start a pivotal study with MZE829 as soon as possible. And so we've already started preparations for that. We're working on the trial design, the scope, et cetera, with more details to come once we're ready. But certainly there'll be a number of factors that influence that, clearly the data that we're continuing to generate with Horizon, as well as other learnings that might come from the field, as well as learnings from the Parasol Initiative, which is a public-private initiative aimed at accelerating drug development in kidney diseases. And they've got some important work going on right now in AMKD, and we'll certainly learn from their recommendations over time and incorporate them into our trial design as appropriate.

I think Parasol's final AMKD meeting is a couple of weeks away. So what are the key questions you think that that analysis could answer? Could they provide recommendations on endpoints at this point, on endpoints for accelerated or final approval? Maybe by, could they look at, provide recommendations by AMKD subpopulations, diabetic versus non-diabetic, and also on time points for measuring protein area versus EGFR?

And so I don't know any specifics about what they're going to show, but we're certainly looking forward to their recommendations. In other areas, such as FSGS, they've worked on defining surrogate endpoints with the goal of, again, accelerating clinical development in kidney diseases, accelerating approvals. And so we'll see what they recommend in AMKD, which is a current project focus for them. And as I mentioned, we'll incorporate them into our own trial design as appropriate once we learn the recommendations.

I know you're involved, Vertex, and I think AstraZeneca are also sponsors of Parasol, the AMKD initiative. Can you discuss how closely the FDA was involved in this initiative and providing feedback to Parasol during their work in AMKD?

Yeah, I know they're involved to some extent, but I don't know all the details behind that.

Okay. And since the sponsors have been involved, did Parasol have access to the full clinical data from sponsors, including from Horizon and also Vertex's Phase II trial, to understand treatment effects in AMKD?

Yeah, it's a good question, unfortunately. I don't know the answer. Okay.

So how are you thinking about Vertex's phase three amplitude trial as read-through to your program? How do you think the mix of FSGS patients in the trial could influence their ability to hit on EGFR slope effect at one year?

It's probably best if Vertex speaks to their study and potential expectations there. We're just really laser-focused on completing enrollment with the Horizon study and showing the results from that study ended this year, early next year, and getting our pivotal trial up and running. You know, over the last four and a half years or so, there have been two clinical readouts in AMKD ourselves and the other sponsor, and the number of patients involved in those readouts has been a grand total of 25, I believe. And so over the next 12 to 16 months, that's going to dramatically change with readouts from ourselves and other sponsors in the field. That combined with the recommendations from Parasol, it's an exciting time to be following AMKD, and I think we're going to all learn quite a bit.

How are you thinking about proportion of FSGS patients in your own pivotal trial? Like, are you planning to cap the number of FSGS patients, you know, to ensure that the trial demonstrates efficacy in a broader population? And how important might that be for getting a broad label in AMKD?

Yeah, more to come on that. We're still working on the trial design. And, again, are really focused on generating, completing enrollment in Horizon. And we're going to have 10 to 15 patients' worth of data in FSGS, as well as 10 to 15 patients in each of the other two segments, patients with diabetes and without diabetes. And we'll use that to inform the trial design for the Pivotal program.

I know we're waiting for parasols analysis, which is, you know, coming shortly. But just curious what your base case is for phase three trial design. Would you approach, you know, something similar to how Vertex designed their trial with a one-year accelerated approval endpoint? Is that how you would approach the FDA for plans for phase three?

I think that's a reasonable starting point or a base case assumption. But this is something that's this is a topic that's under active discussion within the company. And again, we're going to learn so much just even in the coming months where there's from Parasol or other readouts. and most importantly, our own horizon data. But certainly we're going to start or move as quickly as you can in the areas that are most de-risked. And with the horizon data that we released in March, we had our strongest data in FSGS for patients without diabetes, but we're still very much looking across the whole spectrum of patients.

I wanted to discuss diabetic aim, Katie. So the data from Horizon was somewhat mixed, although you did see good responses in two of the five evaluable patients. Considering that mixed data from Horizon, does that make you think that the mechanism works in a subset of diabetes patients, or do you still think APL1 inhibition can work broadly in diabetic AMKD?

Yeah, to your point, that data that we showed back in March in the five diabetic patients. It was promising, but early, and it's probably too early to conclude one way or another. We were certainly very happy to see the responses that we saw in those two patients who achieved per an earlier reductions of 35% and 47%. And recall with not only those patients, but all the patients on the horizon study, there's an eight-week lead-in period where they need to be on stable standard of care background medicines for eight weeks leading into the treatment with MZE829, and that was purposely put in there to try to minimize the noise and try to isolate the effects of MZE829. So the reductions that we saw with all the patients, and specifically those two diabetes patients, is on top of the standard of care treatments that they're undergoing. And we know with AMKD patients who have diabetes, they also, like other AMKD patients accelerate quicker with their CKD and in a more aggressive way. So you would not expect all things being equal, even on their standard of care treatments, for their proteinuria to stay the same. They are still increasing for all intents and purposes and need an additional treatment option. So to see the reductions that we saw, that was incredibly encouraging and promising to see. And as we dose more patients, we'll learn more about whether or not it will be appropriate to enrich future studies or select patients based on biomarkers or other means.

Given some of those factors that you talked about, would you say efficacy bar in diabetic AMKD may be lower than in non-diabetic AMKD? How do you see the benchmark for this population?

The 30% threshold for UACR reduction, that's the one that keeps coming up in conversations with KOLs and practitioners. And again, that's the one that's been linked to improved renal outcomes in adjacent kidney areas. So that's the one hard number we can point to. But it's certainly been suggested to us by KOLs and practitioners that in diabetes, perhaps something lower might be more appropriate. I think this is a discussion that will be informed quite a bit with clinical data over the coming months. And so we'll see what we have and come back to it at that time.

And, you know, how would you weigh average UACR reduction in this population versus percentage of responders and depth of response and diabetic AMKD?

I think all three of those metrics are important, especially when you're talking about a small subset of patients, as we're talking about here. So we're going to be looking at all of that.

Mm-hmm. And then, you know, would you expect more UACR variability in diabetic AMKD? It seems like in some of the trials, for example, with veneranone, there was high variability even in terms of responders on the placebo arm itself. I think a quarter of patients reached that 30% benchmark in UICR reduction. Is this due to high background treatment use in this population? And, you know, would you just expect higher variability in diabetic versus non-diabetic?

I think it's too early to conclude at this point. You know, let us generate the 10 to 15 patients' worth of data in diabetes well in each of the other segments, and then maybe come back to that. But probably too early to say conclusively at this point.

If you do see modest average reduction in diabetic AMKD, but much deeper reduction in a subset of patients, would that be enough to move into a pivotal program in diabetics? Or would you want a way to identify potential responders prospectively before moving into a pivotal study?

I think everything's on the table, and it's going to depend on the actual data. We know that there are no approved treatment options right now for these diabetic patients who have AMKD as well as all patients who suffer from AMKD. And we know that KOLs were really intrigued by even just having two responders among our diabetic patient segments. So, you know, we'll look at the data and make an appropriate decision from there.

I think you've talked about enrichment strategies for diabetic AMKD. You're exploring biomarkers. Can you talk about some of what you're exploring? You know, is it baseline UACR, disease duration, severity of disease, background therapies? Like, what are potential enrichment strategies for this population?

Yeah, it's all of the above, biomarkers, et cetera. I think everything's on the table, and we're looking at the data we've already generated, but we're also going to look at the data that we're generating currently to make some decisions going forward on potential enrichment.

And as far as trial enrollment for the non-diabetic trial that you're starting, the phase three, you know, there's no treatment for MKD, so APL-1 screening has been limited in the past, which explains why Vertex's phase three trial took more than four years, I think, to enroll. Would you expect accelerated enrollment relative to that timeline in your trial because of improved or improving screening rates?

You know, I think awareness is a huge lever with respect to enrollment in AMKD studies, and that awareness is growing exponentially among healthcare practitioners as well as patients. Just given the level of involvement, there's a small but really dedicated group of sponsors in this field, a growing number of patient groups, healthcare practitioners who are starting to focus on the field, and so that will only help enrollment in future studies. And so I would imagine drug development over the coming years will look very different than drug development over the last several years from an enrollment and execution perspective.

All right. Let's talk about MZ782 and PKUs. So the Phase II CIFR trial is now underway, and you're planning to report data in 2027. Can you discuss how you selected the dosing regimens of 120-meg and 240-meg BID, what you saw at those doses in phase one, and why you included the BH4 combination arm in the phase two?

Yeah, happy to. So as I mentioned at the outset, MZE782 is an inhibitor of SLC6A19. That's a transporter of neutral amino acids. And we just announced last month the initiation of a phase two study in CKD. And that's on the heels of some data that we released back in September of 2025. That was a phase one randomized double blind placebo controlled study in roughly over 100 healthy volunteers. And we tested 782 to look at safety and tolerability, but also to measure the level of fee excretion in the urine. and again, these were healthy volunteers. And we had set a threshold there of a roughly tenfold increase in urinary fee excretion. And not only was 7A2 well tolerated, but we exceeded our expectations with respect to urinary fee excretion. We achieved up to over 40-fold increases in urinary fee excretion at some of our top doses. And so we were really pleased with the results, and we started to accelerate our development and move as quickly as we could to a Phase II program, which we just initiated. So we're testing three cohorts. We're going to be in three cohorts with this Phase II PKU study, 120 milligrams BID, 240 milligrams BID, and then 240 milligrams BID in combination with any BH4 agent. So we selected the doses using not only our preclinical work, but looking at the phase one healthy volunteers data in which we saw 240 milligrams BID in particular, regardless of being fed or fasted, no food effect was very effective and achieved a 42 fold increase in urinary fee excretion. So that certainly informed our dose selection for the phase two. And then with respect to adding in combination with a BH4 agent, ultimately in a phase two study, we're going to be looking at the level of plasma fee reduction and see how that urinary fee excretion transitions or translates into plasma fee reductions. And a number of patients, ultimately what the holy grail is, is to get to a point where you can start to liberalize the onerous medical diet that many of these patients are on. These patients can only have a certain amount of protein every day, the equivalent of two eggs, which is fairly onerous, and you need to get them below a certain threshold in order to start to liberalize that diet. So we may be able to achieve that as a monotherapy with MZ782, but for some patients it's possible that dosing it in combination with the BH4 agent may be that additional boost that's needed. And So we wanted to test that out in phase two before we move on to a broader phase three program.

Got it. And with 782, you saw greater urinary fee effects in healthy volunteers compared to Otsuka's compound. How confident are you that that greater excretion can translate into deeper plasma fee reduction in PKU patients and, you know, best in class profile? well.

That's exactly what we're going to test with this phase two studies to see how well that urinary fee excretion result translates into plasma fee reduction. But certainly from our pre-clinical modeling, we certainly believe there's significant potential there. But there's additional sponsor in this field that is also developing an inhibitor of SLC6A19 in PKU patients. They're currently in a phase three study, but we have the benefit of being able to see how their urinary fee excretion levels in their phase one healthy volunteer study translated over in their phase two study with respect to plasma fee reduction. And that certainly gives us confidence that this MOA can translate very nicely in terms of phase two. But that's what we're testing in our own phase two PKU study.

Beyond plasma fee lowering, what are clinical endpoints that we can expect from the trials such as diet liberalization to understand the value of these treatments? You know, where is there still room to improve over the current treatments?

With this phase two study, we're going to be primarily focused certainly on safety and tolerability, but specifically on plasma fee reduction. As we get more into a phase three program will look at diet liberalization. But ultimately, that's going to be what matters for patients and being able to get below certain thresholds.

The CKD trial is planned to start in, I think, the first half of 2017.

First half of next year, that's right.

Yeah. Is there anything gating before starting that study?

No, everything's on track there. As a company, we've certainly prioritized development of 829 and AMKD patients and getting our MZE782 phase two study underway in PKU. But we're putting the finishing touches on trial design and whatnot, normal course items to get the CKD study up and running, the phase two study in the first half of next year.

I guess in the remaining few seconds here, with your current cash runway into 2029, What are the major development milestones you expect to be funded across these programs?

So we ended last quarter with $495 million of cash and investments, and that gives us runway into 2029 and allows us to turn over a number of the data cards that we've been talking about. So it certainly allows us to complete Horizon enrollment and show that data ended this year early next year. It allows us to run the Phase II MZE782 program in PKU and show top-line data in that in 2027, and to get to initiation of the Phase II CKD study and show top-line results from that program as well. So we're funded through a number of these data catalysts over the next several years and into 2029.

All right. Well, we're out of time, so I'll stop there. But thank you so much for being here. Thank you, Son.