All right, so welcome to this next fireside chat with Relay Therapeutics. I'm very pleased to welcome Sanjiv Patel, CEO, and Don Bergstrom, Head of R&D, as well as our friendly representatives from the IR team. And yeah, I'll just jump right into Q&A. Again, my name is Michael Schmidt, Senior Biotech Analyst with Guggenheim. Starting out with your main value driver, RLY-2608, which is your pan-mutant selective PI3K inhibitor, where you've recently disclosed some very interesting interim clinical data from a phase II study with fulvestrant second-line breast cancer patients. And so just remind us of the key takeaways from that interim analysis?
Thanks, Michael. Thanks for the invitation. We appreciate just having to take the Uber over here, so that was very helpful, not having to go to New York. The data that we showed earlier this fall was in hormone receptor positive, HER2- negative, second-line metastatic breast cancer. There are about 13,000 PI3K alpha mutated patients. And they've been underserved chronically now because the agents that have been approved are non-selective, and they're plagued with toxicities of rash, diarrhea, hyperglycemia, and specifically grade 3 hyperglycemia, where these patients require IV insulin. And so our premise over the last five plus years is if we could create a mutant selective inhibitor, we could dial out some of these toxicities, keep patients on drug, and that will translate into better efficacy. And that's hopefully exactly what we showed earlier this fall.
In a heavily pretreated set of patients, we showed a very tolerable safety profile in 64 patients, where we only saw one grade 3 hyperglycemia, which is much lower than any of the previously kind of approved agents, and that translated into a response rate of 33%, which is, again, much greater than what we've seen in the approved agents, and a median PFS of 9.2 months, which is much greater than the five and a half months that's comparable in CDK4/6 pretreated patients that we see either in the real-world evidence in alpelisib or in capivasertib. And so we feel that we have achieved what we set out to achieve, which is create a mutant selective inhibitor, better tolerability profile, translates into better efficacy, and we'll move now rapidly towards engaging with the agency and starting a pivotal trial.
Great. And yeah, so based on the phase II data, as you just mentioned, you've announced potential plans for a randomized phase III study in second-line breast cancer patients, randomized against capivasertib, which is an AKT inhibitor, in non-PTEN AKT patients. And so yeah, remind us again, what are your expectations for capi in the setting, and where patients will likely be perhaps less heavily pretreated than in your initial phase II study?
Yeah, I mean, capivasertib, as you know, is rapidly becoming the standard of care in this field. We saw the sales figures from AstraZeneca yesterday. They have $150 million in their third quarter, so it's exactly what we are hearing in the market from doctors, that this is the best solution for patients, and so maybe, Don, you can just talk about the trial.
Yeah, so you know based on the uptake we're seeing of capivasertib, we are proposing randomizing against capivasertib. So we can look back at the CAPItello-291 trial that led to the approval of capivasertib. And in that trial, the patients were largely second line, so only about 20% of patients were third line or later. About 30% of patients hadn't seen a prior CDK4/6 inhibitor, and 0% of patients had seen a prior SERD. Everybody was fulvestrant naive. You compare that to our ReDiscover patient population, where we've got about 40% of patients who are third line or later, 52% of patients have seen a prior SERD. Everybody's seen a prior CDK4/6.
So as we start looking into the CAPItello-291 trial and look at some of the subgroup analyses that have been presented by AstraZeneca, one subgroup analysis was in patients who had seen a prior CDK4/6 inhibitor, so the 70% of patients who had seen that, the median PFS of capivasertib was five and a half months. Another subgroup analysis they did was patients who had a PIK3CA mutation as opposed to patients who had a PTEN or AKT mutation. And those patients with just a PIK3CA mutation, median PFS was 5.6 months. And that included patients who were CDK4/6 naive. So I think when we put that all together, our estimate for our patient population, which would be CDK4/6 pretreated PIK3CA mutant, would be about five and a half months PFS for capivasertib.
And I think when we look at our ReDiscover population that's more heavily pretreated, where you see over half of patients with prior exposure to a SERD, the fact that we're seeing 9.2 months gives us confidence that even in our more heavily pretreated population, we're seeing a 70% increase in PFS. And we feel as we get into less heavily pretreated patients, it could continue to be favorable for 2608.
Yeah. And then obviously one key advantage of your drug is the fact that it doesn't perhaps impact glucose metabolism in the same way that other PI3K kind of inhibitors do. And so yeah, maybe talk a bit about the competitive landscape and what percent of patients are typically excluded from registration studies based on metabolic markers when you think about alpelisib or inavolisib and other PI3K kinds of inhibitors?
Yeah. Yeah, so I think what we've seen Roche do with the development of inavolisib is given the fact that in phase I-B trials, where they looked in patients who were more representative of Western patient population, they were seeing high rates of high-grade hyperglycemia. So in a standard population similar to our ReDiscover population, where you use hemoglobin A1c up to 7, they were seeing over 20% grade 3 hyperglycemia. When they looked specifically at patients who either had elevated BMI or prediabetic hemoglobin A1c, even with metformin prophylaxis, they were seeing over 40% grade 3 hyperglycemia. So Roche, based on that, decided for the registrational trials for inavolisib to use very stringent inclusion criteria that we estimate based on CDC numbers for the diagnosis of diabetes or prediabetes in the U.S. would exclude about 50% of U.S. patients from inclusion in those trials.
I think consistent with that, now that we've seen a label for inavolisib in PIK3CA-positive, HER2-negative breast cancer patients in combination with palbociclib, we see a significant hyperglycemia warning with actually a recommendation to consult with endocrinologists before initiating treatment in patients who may have a risk of developing hyperglycemia. I think you see both very stringent selection of patients and then fairly onerous recommendations for how to manage hyperglycemia in patients.
Right. And then I know you've talked about presenting the phase II data, which I guess was previewed ahead of ESMO, but the full data set will be at the upcoming San Antonio Breast Cancer Symposium. Just remind us what expectations should be for investors for that upcoming update.
Yeah, we expect to present the same data at SABCS that we disclosed in September. So that's our data set going forward.
This was an encore presentation. Any key items that there were still missing in terms of from, I know, the presentation ahead of ESMO was a fairly complete update, but.
Yeah, we don't anticipate doing additional sub-analyses or anything at this point. I think given numbers, if you start slicing and dicing, they're going to get quite small.
Right. Makes sense. Great. And then remind us, Don, perhaps of your plans for RLY-2608 to advance into the first-line breast cancer study?
Yeah, so earlier this year, we initiated a triplet dose exploration for 2608 with ribociclib, and we chose to combine with ribociclib in this first exploration triplet based on its uptake in the treatment of front-line metastatic patients and the fact that it has a survival benefit in that patient population. Ribociclib has been a challenging drug historically to combine with. Novartis tried to combine alpelisib with ribociclib, and with all the toxicity ribociclib brings to the table and all the toxicity alpelisib brings to the table, they were not able to find a tolerable dose to move forward. We think with the profile we have with 2608, with a much better overall tolerability profile compared to agents like alpelisib, that we'll be able to find a dose, and consistent with that, we've been going through dose escalation.
We did disclose a couple of months ago that we have a favorable drug interaction where ribociclib is actually increasing the blood concentrations of 2608. So we'll use a lower dose of 2608 with ribociclib, but we anticipate we'll get to the same blood concentration that we're looking at in our doublets with fulvestrant. Obviously, once we saw that interaction, we did decrease the dose of 2608 and went carefully to be able to find a dose. So that will be one option as we anticipate moving forward, but we're also anticipating where the field is going. And we've seen, I think, very encouraging data from Pfizer with atirmociclib, their CDK4 selective inhibitor, that dials out a lot of the cytopenias you see with the CDK4/6 class. They've initiated pivotal trials now in a post-CDK4/6 setting. They're focusing on moving earlier with atirmociclib.
As we think about what our options are for the development of 2608 with triplets, I think we've entered into a clinical trial supply agreement with Pfizer. We anticipate by the end of this year, we will initiate the dosing of the triplet of fulvestrant, 2608, and atirmociclib. We think that could be another option of actually going to where the field will be for triplet development, ultimate registrational triplet development of 2608.
Right, right. And then, yeah, I think you talked about also presenting some of the triplet data later this year. Just again, help us set expectations with Ribo for that first part of the study.
I think we'll set very clear expectations that it's early in the trial. We're going to have a small number of patients, and it will be very much focused around safety and what we're trying to dose towards biologically active doses and try and show that we aren't seeing synergistic toxicities. So that will really be the focus of the disclosure that we make.
Is this study being run in first-line patients or in later-stage patients?
No, no. The study at this point is essentially exclusively in patients who have seen at least one prior CDK4/6 inhibitor. It's being run under the umbrella of our ReDiscover protocol, and as such, we're seeing a very similar patient population to whom we reported data out on in the fulvestrant doublet.
Got it. And will you even look at activity measures or just plain safety?
Obviously, eventually we will. But at this point, we're in a patient population that's more similar to, say, the maintained patient population of ribociclib rechallenge. We can eventually look at efficacy, but initially, the question is going to be finding a combinable dose.
Okay, and then I guess when might you ultimately be in a position to make a phase III go decision in first line, and obviously, the atirmociclib cohort is sort of lagging behind. Pfizer is just talking about starting a phase III in first line too. I'm not sure if they have actually committed to doing it next year, very likely they are, but yeah, how do you think about where atirmociclib is in its development stage and then your own in-house data in terms of comparing the two?
Yeah, I mean, I think there are several dependencies. Obviously, getting the experience and getting the dose of the atirmociclib, I think seeing where the space is going, where the field is going. So it's not just a question of where is the field when we launch the phase III study, but where do we think it will be when it ends? And as CDK4/6 inhibitors are moving more and more into the adjuvant setting, I think it will redefine how we think about post-CDK4/6 patients and how we think about the front-line setting. So actually, there could be opportunities to leverage the experience we're getting in patients who have seen CDK4/6 in a metastatic setting to potentially be able to move into a front-line metastatic setting, but in patients who have seen prior CDK4/6 inhibition in the adjuvant setting. So that's going to be something we'll have to monitor.
But I think with the recent approval of Ribo in that patient population, we likely will see a greater and greater proportion of front-line metastatic patients actually being CDK4/6 experienced.
Right. Okay. And then the PI3K space is suddenly becoming, I wouldn't say crowded, but busier. So there's a few other drugs in development now. One interesting one is Celcuity, gedatolisib, in a VIKTORIA-1 study, which we'll read out next year, obviously doing the exact opposite of what you guys are doing in terms of selectivity for that drug. But it will have data next year. And then there's other mutant selective inhibitors from a few other companies. And so what is your view how 2608 is positioned competitively in this category at this point in time?
Yeah, I think we feel very confident that there is a significant place for it. I think some of the agents that you talked about are closer to approval. They obviously have to get over the not inconsiderable hurdle of a successful pivotal trial. I think the agents that you talked about, inavolisib, and I can never pronounce it, but I think you know the Celcuity agent, are both non-selective. And so they come with the significant challenges around not being mutant selective that we've talked about, the tolerability issues. And then obviously, the Celcuity agent comes with being administered IV. And that's a very kind of significant barrier to get over in what is, to all intents and purposes, now a chronic disease that's orally treated.
And so we imagine that even if those two agents do get over the not inconsiderable hurdle of a positive trial, there's still a very clear space for a mutant selective inhibitor with a clear safety and tolerability profile that then will translate into efficacy. In terms of the later mutant selective agents, I think that the data is still very early on those. And many of them have just entered the clinic. And I think provides us with a window that we have now, which is we raised money earlier this fall. We've got a team now in place. We're moving at full speed towards a pivotal trial. And we want to be in the market as rapidly as possible and have that first move advantage that will be open to us being the first mutant selective agent.
Right, great. And so just looking forward to some more disclosures then in breast cancer later this year. And then perhaps switching over to vascular malformations, which is another opportunity for 2608. And I know it's a much bigger patient pool that could be addressed with a PI3K alpha inhibitor. So maybe, Don, just talk about what percentage of patients would actually be treatment candidates for systemic therapy in this study.
I think we hand that one to Megan.
Yeah, so vascular malformations is an umbrella term for a variety of conditions that affect the development of blood and lymphatic vessels. And similar to oncology, PI3K alpha mutations are drivers for a lot of them. So across a variety of subtypes, there's about 170,000 patients in the U.S. And of those, we'd estimate about 25%-40% of them would be seeking systemic therapy. There's only one approved systemic therapy today. And it's only for a small subset of those patients, and it's alpelisib, which was approved through an accelerated approval compassionate use chart review, but is hindered by a lot of the same challenges of non-selectivity that we see in oncology. And then across the other subtypes, there are no approved treatments today. So our team's moving very quickly to start into the clinic the first quarter of next year.
We're also hearing a lot of excitement from the community for the potential for a new treatment option.
Right. And can you talk a bit about your planned POC study and what patients are enrolling and what are you trying to demonstrate?
Yeah, so we are moving into this indication with 2608 because we're able to leverage the fact that we have over 200 oncology patients' worth of experience with 2608 and really understand the dose range for use of 2608, which will allow us, we think, to go into vascular malformations with a somewhat abbreviated dose finding and to be able to look initially across a few different doses of 2608 to really be able to characterize the risk-benefit profile of 2608 in the vascular malformations patient population. We anticipate the initial patients who will come on our trial will likely be PROS patients and lymphatic malformations patients, PROS patients, because that's where alpelisib has the accelerated approval and where we'll be able to benchmark against known data, lymphatic malformations patients, because 80% of those patients have mutations in PIK3CA, and it's a fairly large patient population.
But the goal really would be to look across those doses, characterize the safety profile of 2608, characterize the efficacy profile as measured by volumetric CT, which is the endpoint that Novartis used, and be able to look at some of the other patient-reported outcomes. I think as we get this trial up and running, we'll be able to provide some more details on the design of the trial and on timelines for when we'd expect data.
Right. Okay, then maybe switching gears, a quick question on your FGFR2 inhibitor, RLY-4008, which I know you've slated as a licensing candidate, perhaps. But I know you had some updated data recently at the Triple Meeting. And just help us understand what and you also have FDA feedback on path to market, et cetera. So just help us how that perhaps supports this product's profile and perhaps opportunity to partner.
Yeah, I think over the last few years, we've shown very strong data in FGFR2 fusion-driven cholangiocarcinoma. And obviously, that pivotal trial is now coming towards the end. And so we should have a nice, robust data package there. And I think the data that we showed at the recent Triple Meeting showed that we had extensive activity in FGFR2-driven fusions outside of cholangiocarcinoma, both in terms of response rate and duration of response. We feel very positive about it. So we think there's a larger opportunity that encompasses multiple tumor types. The feedback we have from the FDA is definitely move forward with filing the NDA for cholangiocarcinoma and then subsequently file a supplementary NDA for the tumor types outside. And so that is obviously now a nice, robust data package for us to engage with parties around both completing that and then commercializing it.
Can you comment on what types of arrangements you're considering for the program?
Yeah, absolutely. I mean, as we've talked about for the majority of this, we have a robust set of catalysts in front of us in breast cancer. We've talked a lot about the second-line trial, how we would look at triplets first line. Obviously, I think we have a portfolio that comes behind it around NRAS and Fabry. And so I think our primary goal here is to preserve our capital that we now have to execute those things. So probably the most kind of likely outcome for us would be to look for a global out license.
Okay. And then, yeah, maybe just touching on some of your new molecules. Yeah, Fabry disease is obviously, I think, well-understood condition. There's approved therapies, ERT, and then Galafold on the market for some time now. Just remind us, where do you see opportunity, perhaps relative to Galafold with another oral agent in that market?
Yeah, so as you know, Fabry disease is a well-defined condition leading to inactivity in alpha-Gal A, leading to buildup of Gb3 in various tissues, lung, heart, kidney that has pretty devastating consequences. The approved agents today are ERT or Amicus's Galafold, which is an inhibitory chaperone. It covers about 40% of the known mutations. But it is inhibitory. So the opportunity really states is clear for us, which is, could you go for a much broader range of the mutations that are known? And could you target those kind of so-called non-amenable 60% of mutations? And then secondly, could you remove this challenge of being inhibitory? So Amicus's molecule inhibits the very thing that it's trying to activate.
And so if you could create a molecule that was non-inhibitory, therefore could get deeper activation, you could both treat the 40% of mutations that Amicus is able to address better, and then you could treat the 60% of patients that they can't address. And then I think also more interestingly, you could combine with ERT and not inhibit that either. So I think we think there is a very well-described market opportunity for us. And obviously, Galafold sales is moving towards $500 million at the moment. So it's just not insignificant.
What is the approach to overcome inhibition, technically speaking? What is the approach to, or the angle to, be able to address the non-amenable mutations?
Yeah, I mean, the non-amenable mutations, most of them are rescuable, right? But Galafold is not able to get sufficient chaperoning for those to actually be rescued sufficiently to result in clinical benefit for patients. We think with a non-inhibitory chaperone approach, with a better chaperone that we've been able to discover, we should be able to take these so-called non-amenable mutations to be able to stabilize them and activate them to the level that we'll need to be able to see benefit in the clinic. Now, the assay that's used to decide who's amenable and who's non-amenable in the Galafold label is an in vitro assay. We've been able to essentially replicate that assay and look at our compounds head-to-head against Galafold.
What we're able to show is that across the panel, both of amenable mutations and non-amenable mutations, our agents are able to lead to much greater activation of alpha-Gal A than you're able to see with Galafold.
How predictive are, I guess, or how translatable are preclinical data in this setting?
Yeah, well, I mean, obviously, we are looking directly at the causative mutation of the disease, and we already know from the experience of the ERT and Galafold that if you can stabilize this or replace it, you're going to see that translate into disease modification. Given the fact that this in vitro assay is actually in the Galafold label and it's used to decide who's going to get treated, I think that validates that as having some clinical predictiveness, so the fact that we're able across these assays to see meaningfully superior efficacy of our compounds gives us confidence as we get into the clinic.
Right. And so this is going into phase one next year, right? So maybe then last question just about your other new agent, an NRAS selective inhibitor. So how should we think about the commercial opportunity for that product candidate?
There are about 28,000 NRAS-driven mutated solid tumors in the U.S. each year, and so we believe that they're spread across tumor types such as melanoma, lung, colorectal, so I think there's a significant opportunity. Current treatments are non-selective, toxic, and so it's a very clear precision medicine opportunity. If you could dial out the toxicities, you could get better efficacy, and that should translate into better solutions for these 28,000 patients.
Okay, great. Well, thank you so much, Don, Sanjiv, and Megan. With that, we're at time. So appreciate you being here today.
Thank you. Thanks very much.
Thank you.