Perfect. Welcome to our next session of our Wells Fargo Healthcare Conference. I'm Eva Fortea, one of the biotech analysts here, and we have with us Imogen, Chief Operating Officer, and Pete, Chief Corporate Development Officer of Relay Therapeutics. Thanks so much.
Thank you for having us. Yeah, great to be here.
Awesome. Maybe we can start with kind of lay of the land, past 12 months, next 12 months. What's up for Relay?
Yeah. Thanks again for having us. Great to be here. Great way to kick off the fall season. Over the next 12 months, coming off a very exciting first nine months of the year for us, made a number of disclosures across both vascular anomalies and breast cancer.
We're able to fortify the balance sheet. Now as we look forward, we're really moving into the next leg of growth here as we start preparing to become a commercial company. I think a lot of that will be underpinned by additional disclosures and clarity on the regulatory path forward in both breast cancer and VAs. Specifically, we've guided to regulatory updates for both frontline breast cancer, a phase III trial, which we intend to start early next year.
We will go have a regulatory interaction before the end of this year and report back out on how that goes. Similarly, with vascular anomalies, that program is going quite well and quite fast from a development perspective. We will have a conversation with the agency on a potential accelerated approval path before the end of the year and report back out on that.
From a data perspective, we will share additional vascular anomalies data before the end of the year, and we will share additional phase I/II triplet data in the first half of next year. So continues to be a very robust set of months for us before we get to the end of the year. Then moving our way quickly towards registrational readouts in both breast cancer and vascular anomalies.
Nice. So lots of stuff to talk about today. Maybe we can start with the breast cancer side of the story. Maybe you can give us a quick overview of the data we've seen so far for zovegalisib in breast cancer and what are key differentiators, especially in this very competitive landscape?
Yeah. So we'll start with the second line because that's where we've got our current phase III running. The goal with zovegalisib when we set out was really to try and design a drug that was very specific and did not have the limitations of the generations that came before it, and so avoided some of the off-target toxicity. We went after the second line breast cancer population first because that was the biggest unmet need and the place where it made sense to start. What we were able to show, and Pete alluded to this earlier this year, we showed fed data.
We showed that we had an 11-month PFS, which was really exciting because it avoided some of the off-target toxicity, had a much cleaner profile, and it delivered against this promise of, if you could have a better tolerability profile, you should be able to extend the efficacy. That compares really nicely to capivasertib's five and a half months in a post- CDK4/6 setting. So we're really excited about that, and that's what gave us confidence to launch our second line phase III, which we started last year.
Then earlier this year, we showed data from a median third-line population in the triplet setting also that Pete alluded to, and that looked really compelling both from an early efficacy perspective, we showed an ORR of 44% there. But more importantly, we wanted to show that from a combinability perspective, it would be a really nice tolerable regimen for patients in the front line and could compete well with the existing standard of care.
Got it. So maybe starting with the ReDiscover-2 phase II
Yeah
study, just in terms of enrollment, is it tracking line expectations and how should we think about timeline to read out?
Yeah. I think this is probably one of the most important priorities for us as a company right now. And it's going really well. It's a very compelling study for physicians and patients, given that you are taking zovegalisib plus fulvestrant versus capivasertib, the market leader right now, versus fulvestrant. People are very excited about it, and enrollment's going well. As Pete mentioned, we'll guide before the end of the year as to when last patient in will be. But right now, we're feeling really good about it, and we think that we will do very well compared to benchmarks that we've seen for similarly sized studies.
Got it. Should we expect any impact on this study given the recent focus on the first line?
No, not at all. I think for us, it's one of the key priorities, as I mentioned. It's really important in terms of the patient population. There's a lot of excitement, we will have a completely separate team working on standing up the frontline study.
Got it.
It's kind of nice. As this study will reach full enrollment, you're then just collecting events. In contrast, we'll have the frontline phase III trial getting up, we can have the team shifting and focusing on enrollment into that study on a global basis. We'll use distinct teams internally, but we'll leverage a lot of the global infrastructure we put in place to be able to execute against ReDiscover-2.
Got it. You mentioned capi as the control arm, but we've seen recent data from one of your competitors, gedatolisib, that looks very compelling. Do you think this sets a new bar, or are you still thinking of capi as the bar to beat?
We still see capi as the bar to beat. I think, gedatolisib came out with an 11-month PFS in their data in PIK3CA-mutated patients at ASCO. We think that when you look at the weekly IV, that's just such, it's going to be so onerous for patients in this setting. It's going to be a very hard, it's tough, especially as you think about our own efficacy sitting at 11 months. If you compare a weekly IV versus an oral option, people are going to want to go for the oral option. I think it's an exciting drug in the wild-type setting where there are limited options, but for PIK3CA-mutated patients, I think we can do better.
I think the dynamic that happened between Piqray and Truqap in this setting tell us everything we need to know to that question. You had Piqray with about a six-month median PFS based off of the phase III trial that led to approval but was encumbered by safety and tolerability issues, namely grade three plus hyperglycemia, but also rash, stomatitis, GI toxicities.
Truqap actually had a nominally smaller median progression-free survival of five and a half months, but just had a slight trading off in side effects with numerically lower grade three hyperglycemia and has now become the market leader in the space. I think it really tells you that the safety tolerability profile is paramount for these patients, especially in the second-line setting. I think that direct comparison between us and really any other regimen that's available today favors zovegalisib quite a bit here.
Got it. Can you just remind us a little bit on how zovegalisib's tox profile compares, and how are you thinking about the market opportunity in terms of the second line?
Yeah. On the tolerability profile, we've seen improvements across the board. We have no stomatitis rash to speak of in our AE table for the second-line patients. Our grade three hyperglycemia rate remains in the low single digits across over 100 patients treated at this exposure. In the context of whether it's gedatolisib, or capivasertib, or alpelisib, you see that in case of capivasertib, there's a high degree of stomatitis, rash, and diarrhea.
In the case of gedatolisib, you need three or four different concomitant therapies just to keep the grade three stomatitis in check, and the rash. I don't think we're seeing, one, we don't need concomitant therapies to prophylactically treat these side effects. And two, the overall absolute rates that we're seeing are much more favorable compared to these other agents.
Got it. Maybe talking a little bit about the first-line opportunity. Earlier this year, you selected atirmociclib as the CDK4 of choice for the combination. Where are you at in terms of regulatory conversations, and how do you think about the path forward in the first line?
Yeah. As Pete mentioned, we've guided towards coming back with regulatory feedback before the end of the year. That being said, what we're envisioning is a reasonably straightforward design. Our goal is to go after the endocrine-sensitive patient population with a study that largely emulates Pfizer's FourLight-3 study of atirmociclib in the same patient population.
We went after the endocrine-sensitive population because it's the larger portion of the frontline setting. We haven't ruled out going into endocrine-resistant patients, but I think as we thought about what made sense for our initial move, that was the most compelling thing to do. We're still thinking about whether or not to add a small contribution of parts arm to the study and whether or not we add an atirmociclib plus AI to help us ex-U.S.
Within the U.S., the FDA has been pretty clear that we could reference FourLight-3, but I think outside of the U.S., it will be helpful to have at least a small contribution of parts arm. Our goal is to set that study up, to get it designed and approved, and then hopefully be ready to launch in the first half of next year, early part of 2027.
Got it. In terms of durability, follow-up was quite short from the one that we've seen so far for the triplet, but how should we be thinking about next update? How much data on durability are we going to get?
Yeah. I think the biggest thing that we're looking at in the phase I/II data in the initial update and in subsequent updates will continue to be the safety and tolerability of the regimen. That is going to be the best indicator of eventual durability because we already have good analog data from inavolisib in the frontline.
The INAVO120 study was inavolisib, a non-selective inhibitor, plus palbociclib, plus fulvestrant versus palbociclib and fulvestrant in endocrine-resistant patients. In that trial, they had to have a very favorable metabolic profile of the patients in order to keep the hyperglycemia at bay. But the efficacy was quite profound. You saw a doubling of the PFS in that patient population, which tells you that when you add a pathway inhibitor to these patients in the frontline, you're going to get a dramatic prolonging of progression-free survival.
The real key question is, when you contemplate the frontline endocrine-sensitive population, where you could be treating these patients for upwards of three years, can you put together a regimen that is tolerable enough to stay on for three years? In the phase I/II data that we continue to generate, of course, we'll continue to report on response rate, and eventually, if we get enough data maturity, we can look at progression-free survival in that setting. But the key metric we'll continue to look at is the safety and tolerability of that regimen.
Got it. Are there any safety events that we should expect to accumulate over time versus seeing them at the beginning?
I think, when you treat any patient population over the course of three years, especially one where it is a metastatic cancer population, and the median age is somewhere in the neighborhood of 60 years old, you are going to see accumulation of certain AEs. I think most of the AEs that we have seen thus far with zovegalisib do tend to be early on, tend to be low grade. Some of them are reversible. They all appear to be exceedingly manageable.
I think in the context of combining with a CDK pathway inhibitor, the one addition you are definitely going to see is neutropenia. That is a class AE that we have seen across the CDK class of therapies. It looks like in the previous atirmociclib data that the grade three/four neutropenia tends to be a little bit less than what you see with palbociclib, abemaciclib, or ribociclib. The general tolerability of that agent versus the incumbent CDK4/6s seems to be better than what we have seen historically.
Got it. You also showed in the early data that it seemed like there was higher exposure, like a 2.5-fold increase exposure for zovegalisib. Is there any risk in dose selection for the phase III? How should we be thinking about that?
No, we don't think so. We feel pretty confident in the exposure that we've seen and in the dose that we intend to move forward. We know that we do have an exposure DDI with atirmociclib, but we don't see dramatic differences in interpatient variability or something like that. So it's a very manageable, predictable exposure increase. We feel confident moving forward with our RP3D.
Got it. Maybe in terms of control arm, the landscape is rapidly changing. Do you foresee any unexpected issues with your trial design or anything that could impact the way you're thinking about it?
I think today, as we sit here now, the CDK4/6 plus endocrine therapy is very much the standard of care. This is the best design that we can design from where we sit today. There are a number of CDK4s that are in development. We're seeing the rise of some of the oral SERDs. I think from where we are now, this is the design that makes sense, and this is part of the reason why we wanted to bring a selective PI3K with a selective CDK4 plus the endocrine backbone because we felt that that was going to be one of the most future-proof designs that we could bring to bear.
As we think about the two main questions you're trying to answer there is, will global regulators view it as a regulatory standard of care? I think we can definitively say for the next five years or so that that question's going to remain yes. Then, is it a study that physicians on a global basis will want to enroll to? I think, again, the answer to that question is yes.
You're offering CDK4/6 of choice, along with the standard aromatase inhibitor backbone in the endocrine-sensitive population. The option is to randomize into a treatment arm where, as Imogen said, you're offering the selective approach, which based off of the data we have in hand today and some of the data we've seen from atirmociclib thus far, should be a favorable option against the existing standard of care.
Got it. Can you remind us how much data should we expect in the first half of 2027 as we get more information about this first-line strategy?
Yeah. We continue to enroll in second-line plus patients, with zovegalisib plus atirmociclib plus fulvestrant. We've also enrolled smaller arms to look at zovegalisib plus atirmociclib plus aromatase inhibitor. We are also trying to enroll or are enrolling a smaller portion of patients that are endocrine-resistant with the zovegalisib plus atirmociclib plus fulvestrant regimen. The bulk of the patients we'll see being reported out from the phase I/II will be this median third plus line of therapies, which is why we continue to focus on the safety and tolerability that we're getting out of this data set.
Got it. Is there any reason to believe this third-line data wouldn't translate well to the first line?
We've seen this. The precedence is that it translates favorably. You generally see, if you look at the inavolisib experience and in smaller capivasertib studies, you tend to see response rate improve by 20% to 30% when you go to second or third-line plus patients into the front-line patient population.
Got it. Very helpful. Maybe just last question within the breast cancer space. How are you thinking about prioritizing HER2-positive patients or triple-negative?
From a biological perspective, it is definitely interesting. I think as a company, where we chose to focus first has been HR-positive, HER2-negative metastatic breast cancer, and I think that remains our focus. Of course, we will continue to think about these other opportunities, but for now, we are going to focus here because we see it as the biggest opportunity.
There is also, before we even get into triple-negative, one of the settings we think about in HR-positive, HER2-negative is the adjuvant setting. I think there are future opportunities in HR-positive, HER2-negative disease. You could contemplate Roche Genentech are running combination studies in HER2-positive disease. There is definitely a broader breast cancer opportunity as we continue to think about ways to expand the utility of zovegalisib.
Yeah. Got it. Very helpful. I do not know if there are any questions for the breast cancer side of the story before I move on. No? Okay. Then I am going to move on to the vascular malformation side of the story. Very exciting data that you have shown. Can you put some of the data into context of how does it compare with the other PI3K available to date?
Yeah. We came into vascular anomalies with the same hypothesis we had going into breast cancer. You had good, clear clinical proof of concept with a 20+ year-old non-selective inhibitor in alpelisib, and our hypothesis was that with a selective inhibitor, you could push the dose intensity, get better efficacy, and that could also, because of the selectivity window, come with a better safety profile.
I think the initial data that we disclosed at ISSVA in May really checked all those boxes and then some in certain contexts. We started the study in the middle of 2025, and then presented initial data from the dose-randomized portion of the study in May of this year at ISSVA. We saw across the dose-randomized patients a 60% response rate.
Importantly, some of the early indications for patients that have made it out past multiple MRI scans, we are seeing deepening of benefit for those patients. It is coming with good symptomatic improvement across the board in both investigator- and patient-reported outcomes, and specific when looking at pain, which is a common symptom that a lot of these patients have to deal with.
It compares quite favorably to what we have seen with alpelisib to date, which has a 20%-30% response rate at the labeled dose. Here in our randomized dose finding, we are seeing already a 60% response rate at unoptimized doses at an early time point. I think we are extremely excited about what is to come from a differentiation standpoint for zovegalisib in vascular anomalies.
Got it. How are you thinking about further dose optimizing, and also how should this data translate to the younger cohorts?
Yeah. Earlier this year at ISSVA, we announced that we have started the dose expansion portion of this study. We brought both 300 mg BID and 400 mg once daily into expansion. We are currently prioritizing enrollment in the 400-mg once daily dose. We think that is a real good sweet spot between 100 mg BID, which was extremely well-tolerated and was showing similar efficacy, modestly better efficacy to alpelisib, but dramatically better safety.
At 300 mg twice daily, we saw 100% response rate, and it came with a little bit more tolerability profile, but still better than alpelisib. This 400 mg once daily might end up being a really good balance of safety, tolerability, and efficacy and lead to meaningful differentiation from both safety and efficacy against alpelisib and ultimately sirolimus.
We are very excited about continuing to roll into the expansions and really confirming a go-forward dose. That is all in the 12 and older population. We also started dose finding in the six to 11 population. There it is a more traditional dose escalation study, so we are moving through weight-based dosing there, and the goal is to get to a weight-based dose equivalent of roughly 400 mg once daily exposure.
I think ultimately we can get to similar once daily options in both the 12 and up patients and then in the weight-based dosing in the six to 11s. Then ultimately, we will move even lower into the two to five-year-olds once we identify a go-forward dose in six to 11s.
Got it. Is the physiology and the dynamics of the disease similar in the older patient population or in the younger? Should we be expecting any differences there?
Biology, we do not really expect any difference or impact on the disease. Clearly, you would like to intervene as early in childhood as possible. These lesions tend to grow with the patients over time, so if you can intervene early in childhood, you have the opportunity to maybe stave off disease severity as the child gets older. But we have no reason to believe that the impact there, the disease biology, is any different in younger patients than older patients.
Got it. You previously mentioned some deepening of responses during the data cut in May. We did not have a ton of follow-up. How important is it, durability versus the responses and the quality-of-life measures that you can get, like pain at the beginning?
I think durability is going to be critical to maintain over time. Yes, in a handful of patients, we were able to get out to 24-week scans, so two scans, and we saw in every patient that was in response going into the 24-week scan, we saw a deepening of the response at 24 weeks. We also saw sequential improvement of the symptoms over some of the early time points in the study. We will be critically keeping an eye on the ability to maintain response with these patients. Probably equally as important, if not more important, the symptomatic improvement over time.
Got it. Were there any type of patients that did better than others that you could point out? Also, what was the status of the PIK3 mutations? If I recall correctly, there was a small subset of patients that did not really have a PIK3 mutation there.
Yeah. We reported on many different subtypes of, or different cuts of the data, with the caveat that in these early disclosures, once you get into these subcuts, you get into pretty small ends. We were encouraged to see that regardless of the subset, we are seeing similar activity.
So whether it is PROS or non-PROS disease, whether it is a kinase mutation, non-kinase mutation, prior sirolimus and alpelisib or not, we tend to see very similar outcomes for these patients thus far, which was quite encouraging to see. You highlight specifically there are some patients in the study without an undocumented PI3Kα mutation. The reason for that is we allow that for both PROS and lymphatic malformation patients, because those patients can be diagnosed clinically and do not need a genetic test to direct treatment.
Because PROS patients are 100% PIK3CA mutated, and the lymphatic malformation patients are 80+% PIK3CA mutated, we feel comfortable allowing undocumented patients coming into the study because almost certainly they are going to have a PIK3CA mutation. There are some times, because the mosaicism of the disease, in order to do a genetic test, you have to take a tissue biopsy of the disease, and that's sometimes not feasible in certain patients.
Got it.
Also, in some patients where you do have a biopsy, they may have very low variant allele frequency of the mutation that may be below the lower limit of detection of the diagnostic test being used. So, for instance, our lymphatic malformation patient that we reported on that has the greatest depth of response in our data set today does not have a documented PIK3CA mutation. Clearly, the patient is PIK3CA mutated. It's just the technical limitations of the diagnostic test have not detected a PIK3CA mutation.
Got it. You also mentioned patients that were alpelisib, sirolimus experience. Were this patient intolerant? Had they progressed?
The majority of them came off of sirolimus or alpelisib due to intolerance of some kind, or they've plateaued at a level that was unsatisfactory to the patient and the physician. I think those are the two main reasons we see patients switching to zovegalisib, is the prospects of getting deeper volumetric reduction, better symptomatic improvement, and that coming with a better safety and tolerability profile.
Got it. Just in terms of washout period, did you have a mandate washout period for patients to enter the study?
Yeah. We have a very standard washout period of three half-lives of the drug.
Got it.
That happens within a couple of days of alpelisib. What doesn't happen in those couple of days is complete reconstitution of whatever side effects that those drugs were causing. So we will, because we're being so inclusive in the study, we'll continue to see some of the baggage of having been treated for longer periods of time with sirolimus or alpelisib may kind of trickle over into our study. But those patients have such high unmet medical need, and it will be an important patient population for us that it was worth it from our perspective to make sure we're allowing those patients into the study.
Got it. You mentioned a lymphatic malformation patient before. What was the split between your PROS patients and your lymphatic malformations patients, and have you seen any difference in terms of how they react to the drug?
Yeah, we're currently seeing roughly 70%, 75% PROS patients, 25%, 30% lymphatic malformation patients today in the study. Not surprising to see this split given that the PROS patients tend to be, as a percentage of patients that are severe, they're much higher percentage of those patients are severe patients, as indicated by the fact that they have syndromic disease.
Then that's the indication where you have alpelisib with accelerated approval. But to date, we've seen no difference in performance of zovegalisib in either the PROS or lymphatic malformation patients. As I alluded to, some of our deepest responses have come in the lymphatic malformation patients.
Got it. You mentioned you are going to be talking to the FDA later this year. What are the key points that you need to discuss, and is an accelerated approval path on the table, and how would that look like?
Yeah. The current precedence, there's only one approval in this space. In contrast to what Imogen's tackling on the breast cancer side, on the VA side, it's a very nascent disease area with not a lot of precedents. But the one approval we do have is alpelisib has accelerated approval in PROS based off of a very strange study, which was 37 patients treated under compassionate use, and it was a retrospective chart review of those patients.
Yes, we do believe accelerated approval should be available or could be available to us. What we will propose to the agency is a pooling of PROS and lymphatic malformation patients, and a single arm accelerated approval path to address both of those patients, both patient populations. The question will be if they agree with that or if they would like to see those subgroups split out. What we would also like to get a sense of is the size of that data set required for accelerated approval. Those will be the two key things we're trying to elucidate with the agency.
Got it. Based on the precedent, how are you thinking about potential timing for this accelerated approval? Also, it's a little bit more clear for PROS than it is for lymphatic malformations, but do you think you're going to need the same amount of data, or you could get away with more data for PROS than lymphatic malformations?
I'll refrain from getting too specific there until we actually have a conversation with the agency. They're ultimately the ones that have the most say in that. But what we do know, the current regulatory endpoint is exclusion. It's a volumetric response rate endpoint, exclusion of 15% in the lower bound of the confidence interval for your data set. With the magnitude of benefit we're seeing today, we don't believe it would require that many patients to be able to achieve that outcome in both PROS and lymphatic malformation patients.
Got it. How are you thinking about the opportunity there? How should we be thinking about PROS versus lymphatic malformations? What is next for the vascular anomalies program after this?
Yeah. We think it is a very large opportunity. Just within PROS and lymphatic malformations, there is probably somewhere in the order of magnitude of about 15,000 patients in the U.S. with moderate severe disease that should be addressable with chronic systemic therapy like zovegalisib. As a reference point, if you have Vijoice-like pricing, which is about $36,000 a month, every 2,500 to 3,000 patients would result in a $1 billion in peak sales if you can keep these patients on therapy chronically.
Got it. Okay. What does chronically mean here?
Exactly what it sounds like. You start the patients as early in child as you can and keep them on-
Forever
for life.
Got it. Very helpful. We are out of time. Thanks so much for joining us today.
Thanks.
Thank you for having us.