Where that whole space is moving over the coming years, and how are you thinking about positioning your mutant selective PI3K α compound?
Absolutely, Simon, that's a great question. So I think, as you know, breast cancer is a very dynamic space, and there are many new agents that are currently in clinical trials and on the horizon. In general, we think about the space as moving towards next generation all-oral combinations. And while it's not clear which of these compounds will ultimately win, we believe physicians will always choose to use a targeted agent for PIK3CA- mutated patients if the right one is available. So for RLY-2608, when we think about the positioning of that, we really think about it as being the backbone PI3Kα therapy, which can be combined with other agents like some of these new oral SERDs that we're seeing and/or new next generation CDK agents.
Great. Yeah, maybe digging down in each of them a little bit. How are you viewing the SERD space evolving? Obviously, there's several important updates at the end of the year at SABCS. And then how is that relevant for your asset? Because you're obviously studying it in combination with one of the SERDs.
Yeah. So I think when it comes to the SERD landscape, we think that this shift towards oral agents is really in line with an overall trend in breast cancer treatment towards well-managed long-term palliative care. We've seen that elacestrant came out. It demonstrated benefit over fulvestrant in patients, at least in patients with ESR1 mutations. And we think people are going to be watching really closely some of these upcoming trials to see if they see similar improvement for patients who don't have an ESR1 mutation. When we think about RLY-2608, we really think that the highly selective and combinable profile that we have should be well tolerated in combination with any one of these next generation SERDs. And so I think we'll be watching the data readouts really closely over the next several months to see which one is likely to emerge as the winner.
And that will become important as we think about our go-forward development plan and how we might exchange fulvestrant for one of these next-generation SERDs that we see. So it's a bit early for us to commit to one in particular, but I think as we see some of the data readouts come through, we can think about doing phase two type studies where we can look at these agents in combination with RLY-2608. This would then hopefully build off of the phase III that we're planning, and we can think about those sort of in conjunction together.
And just adding to what Imogen mentioned on the phase III, I think the space is evolving, but it won't have evolved by next year, which is when we're working to start our phase III pivotal trial in the second- line. And so I don't think you're going to see an impact what we use as the SERD in the trial. It would still be fulvestrant, but as Imogen said, then we could investigate some of the oral next generation SERDs in supported phase II trials or something like that as what does emerge as a leader.
Great. Yeah, thank you. Yeah, just for listeners, obviously, we expect EMBER-3 data and ELEVATE data at San Antonio Breast Cancer Symposium. And then in terms of the safety profile, anything that could come out of that conference that will make you think about your combos and also going forward in combination with any of those or any toxicities that would be unexpected from those two new compounds that would maybe synergize poorly with your compound?
I mean, I think what we've seen to date for our compound is quite a combinable safety profile. I don't think we would anticipate seeing anything with any of these next generation oral SERDs that would concern us. But as you can imagine, we're all quite excited to see this data come out, and I think it's eagerly anticipated by a lot of people in the field because it does represent sort of the next generation of options for patients.
Then also CDK4/6, the next piece that's important as a combination for you. How are you looking at that space over the next couple of years as the next generation of molecules comes out here?
Yeah, I think also evolving, which is obviously the key theme here. But we know that CDK4/6 is bringing a significant level of toxicity. While the clinical community has figured out how to manage them, that doesn't mean there's not a desire to dial them out. And so I think you're seeing that in some of the next gen approaches with the CDK4 selectives. Pfizer is most advanced with that. And we have the clinical trial collaboration with them to include atirmociclib as one of our triplet combinations. And we're on track to start that triplet this year. And similar to how RLY-2608 dials out the off-target toxicities of the non-selective PI3K agents, you're seeing the data that we've seen to date from Pfizer's CDK4, they're dialing out some of that CDK6 toxicity.
So this ability to combine two more selective approaches, I think, has great potential for patients going forward.
Yeah, great. And obviously, greater selectivity with less liability leads probably to easier combinability, which is a positive for you here. And then another recent shift was obviously. What do you think is the impact of the recent inavolisib triplet approval in breast cancer here? Does it have any synergies to any of your programs or plans?
Yeah, so I think it provides the data that they presented at SABCS almost a year ago now provide great proof of mechanism showing that the addition of a pathway inhibitor does lead to additional efficacy, which I think had been an outstanding question in the field of could you get any more efficacy? And it shows that you can. However, given the non-selective nature of inavolisib, they had to carefully select their trial population to manage the hyperglycemia liability. And so it excludes a lot of patients. Even in that endocrine resistant frontline population, about half of Western adults are excluded from the trial. The label they got does not include that restriction. However, the guidance in terms of managing hyperglycemia and the label is very similar to the alpelisib guidance in their label. And so we know that the hyperglycemia liability for alpelisib has really limited its use.
And so similarly, we don't expect there to be broad use of the inavolisib triplet despite the fact the label is broader. And so in terms of kind of impact on the space, I think really important proof of mechanism, but also there's a lot of room to improve on it with a selective agent.
All right. So we should be tracking closely that launch and kind of like triangulate your market size. It will be significantly different if your profile holds up. Talking about that, maybe diving into RLY- 2608, your compound in this space. Can you just give us, for the people unfamiliar with the program, could you just give us a quick overview of the settings you're pursuing here?
Yeah, yeah, happy to. So RLY-2608 is our lead PI3K asset. And it's been designed to be selective for the mutant form of PI3K over the wild type form. And the wild type form is where you get a lot of the toxicity that are limiting the class of first gen agents that we were just talking about, inavolisib, alpelisib, and capivasertib. And so by dialing out those off-target toxicities, we're seeing a much better safety profile with RLY-2608. And that's leading to better efficacy. And so we're developing it across a few different areas. From a breast cancer perspective, which is our main and our lead focus right now, we have the d oublet study, which is the data that we shared in September, which I know we'll get to. And then those data are supporting us starting a pivotal trial next year in the second- line patients.
And then we also are investigating its potential in the front line. And so that would be in the triplet combinations that we've talked about with the CDK4/6 or CDK4 in the future.
Great. So yeah, talking about that data, could you just please walk us through the update in September?
Absolutely. So in September, we were very excited to announce our RLY-2608 doublet data. And this showed that we had a 9.2-month median PFS in second-line plus metastatic HR+/HER2- breast cancer patients. This is nearly a four-month improvement relative to capivasertib, which is the regulatory and emerging commercial standard of care, which has a median PFS of 5 in patients who have previously seen a CDK4/6 agent. Similarly, as Megan mentioned, in terms of other agents, Piqray, alpelisib has shown median PFS of 5.6 months in their Cohort C of the BYLieve trial, which is the cohort that matches the most closely with the data that we just showed from the ReDiscover trial. So the data continued to show RLY-2608 selectivity with limited off-target toxicity at the RP2D of 600 mg BID.
We had only 25% of patients experiencing a grade three treatment related AE. We had no Grade 4 or 5 events. We only had one patient who had a G rade 3 hyperglycemia event, and this patient was also receiving steroids, which are known to be an independent risk factor for hyperglycemia, so these data gave us a huge amount of optimism that RLY-2608 will be able to demonstrate really meaningful superiority against capivasertib in a second- line pivotal trial, which we're looking to start in 2025, so next year.
Great, thank you for the overview. And obviously, you already mentioned that it's very important to dig down to the subpopulations here because comparing between trials and how to compare to capivasertib, which population is important. But maybe what can you tell us about your subpopulations at this point? It's still early days, but for example, including ESR1, et cetera, some of these groups.
Yeah, it's a great question. And it's something that we spend a lot of time looking into. The short answer is that we haven't seen, across, in any subpopulation any sort of meaningful deviations from the PFS. So this is true when we look at patients with ESR1 mutations, also true when we look at patients who've received prior SERD treatment. This also includes PIK3CA mutation types. So we have looked at kinase mutations versus non-kinase mutations. And across all of these patients, we see good efficacy signal. Based on the biologic rationale for a mutant selective PI3Kα inhibitor, we have decided to restrict enrollment for our phase III pivotal population to exclude PTEN and AKT co-mutations. These occur downstream of PI3Kα. And so we have seen a reduction in efficacy benefit, as you would expect, for patients with these mutations.
But it only includes 5%-10% of the population that we're talking about. And it's consistent with other studies that are looking at PI3Kα inhibitors that they're all excluding PTEN and AKT co-mutations. So it's not particularly surprising. But in general, we don't see any dramatic differences in subpopulation analysis.
Yeah, and I think that certainly mechanistically makes sense. Can you talk about the lack of impact on glucose metabolism? You already went over it, but what is the tangible meaning of this in respect to the label that we've talked about earlier of inavolisib and certainly alpelisib? What would that mean for patients and I guess use?
I think generally speaking, this is going to mean that the tolerability profile allows you to use this in a much broader range of patients, as evidenced by only having one instance of grade three hyperglycemia in this data set. I think what this means is that you can potentially use this in a broader range of patients, and then also, as we've been discussing, it will be hopefully combinable with other agents, and so the fact that you have this much cleaner profile means that you can use it in other combinations, so I think that's sort of the biggest take here.
I think the other thing just to kind of help ground people is just looking at the grading of hyperglycemia. So you have grade one hyperglycemia, which requires no medical management. And then you have grade two hyperglycemia, which is generally like an oral antihyperglycemic. Once you get into grade three is when you need an endocrine consult and injectable insulin. And so that's really where the focus is, is minimizing that Grade 3 and greater because that's what is really impacting physicians' treatment decisions because that kind of hyperglycemia is not something that oncologists are typically used to having to manage and just makes them more uncomfortable. And so obviously, we want to minimize all grade hyperglycemia, but the real focus is grade three and greater because that's what impacts physicians' treatment decisions.
Frankly, in a Western population, many people are walking around with Grade 1 hyperglycemia after lunch each day anyway.
Yeah, that's certainly me. What are some of the drug-drug interactions that you worry about with this compound? And I mean, obviously, combination is the key here. Can you use the full dose with fulvestrant? And why is that important?
Yeah, so the only interaction that we have seen to date is in September, we disclosed that with the ribo combination, that ribociclib increases the exposure of RLY-2608. And so that means that we've gone back and started dosing lower in the triplet arm. But that we think we'll get to the same exposures just with less RLY-2608. RLY-2608 is not changing the exposures of ribo. It's just a victim of ribo from that perspective. But other than that, we haven't seen any interactions that we'd worry about. And in just the doublet with fulvestrant, we haven't seen anything there either.
Scorpion, one of your competitors, saw a DLT of paresthesia and myalgia and also some grade three, four liver tox. Is there a risk to your program or is that compound specific?
It seems to be compound-specific. I think the paresthesia, the myalgia, even the liver tox aren't things that you've seen either with inavolisib and even with the first-gen agents. They've never been seen before. So it seems like that would suggest that it's probably a molecule-specific type of idiosyncratic tox. They were obviously treating late-line phase I patient population. So you never know kind of what comes along with that. But I think all of us have treated those late-line phase I patient populations, both ourselves as well as the first-gen agents. And no one's seen those toxicities to date. And so I think as we've talked about, combination therapy is going to be the treatment of the future. And so I think we're very pleased that we haven't seen them.
We think that that is good for our potential to combine going forward.
Great. And then maybe zooming out, obviously, that is a very confident target. We've tried for a long time to drug this selectively. Can you please outline, based on your data, how you compare to several clinical stage PI3K programs, including we have a lot of them still only have preclinical data, but we have Celcuity. We get a, I can never pronounce this thing, gedatolisib. OnKure, we're going to have data, I think, at San Antonio Breast Cancer of OKI-219. And then, of course, we have Scorpion's STX-478. And Lilly had one and decided to swap to a follow-on compound. How do you compare? And should we be comparing here or are the differences just very subtle at this point?
Yeah, I think there's a few ways to take it based on just the mechanisms and kind of what they're hitting. So if you look at the OnKure compound as well as the LOXO-783, those are H1047R specific. So just for a subset of the population, their approach is to get very, very selective for that one mutation. I think what we see is that we are selective for the mutant form, as we've talked about, over the wild type form, but not at the levels they're looking at. We are the only ones, as you said, to have shown clinical data. And we've set a pretty high bar, I think, for those patients. The 9.2 months that Imogen mentioned earlier is the highest ever seen in the class. And that's across all mutations. It was even a little bit higher, I think, 10.4 months in the kinase-only patients.
So that's more of that like-for-like patient population. So I think the question is, with setting a bar that high, is that additional selectivity that they're getting actually going to have any clinical meaningfulness? And we had our own H1047R specific program and don't feel like we need it. So haven't advanced it into the clinic at this point. But I think that the emphasis is on them to show that that additional selectivity actually leads to a clinically meaningful difference. So that's something that we'll be looking for there. So that's those agents. Then if you take Scorpion, I think we already talked about the potential idiosyncratic tox that have been seen there. Other than that, I think that their agent looks fairly similar to ours. In our hands, the structure is public. And so we've made it and profiled it. But there's the idiosyncratic tox piece.
Then they're just a couple of years behind us in the clinic. Additionally, we have experience running global trials. We also have the capital to do it. So I think that's where we see the differentiation there. Then the last one was Celcuity's gedatolisib. That one's in a different class from the other two that we've talked about because they're going after both the mutant and the wild type forms in their different studies. Imogen had earlier, though, talked about how the whole field is moving toward all oral regimens. So I think that the fact that they are infused is going to be a significant challenge to adopt. When otherwise you're moving toward all oral regimens, to go backwards to an infused agent seems like it's going to be difficult.
So I think it creates a much higher bar for what they have to show to make it actually competitive commercially.
Great. Looking forward to the triplet safety data, which we may get in December at some point, can you please characterize what we should look for in the triple safety data?
Yeah, so I think with the triplet safety or just this initial work that we're doing in the triplets, the point is to demonstrate combinability and to identify a dose, and so what we're looking for is to show that safety, really, that there's no synergistic toxicity, that it's really each agent bringing their own safety profile, their own known AEs to the table, that they're not being increased by the other agent, and that there's no new toxicities that you're seeing as well.
Great. And what are the key AEs that we should focus on? And is neutropenia one of them?
Yeah, so as I said, from a kind of whole profile perspective, it's that we just want to see some of the parts. We don't want to see new AEs. We don't want to see synergistic toxicity. And so with the PI3K agents, we know that the key toxicities there to watch for are the hyperglycemia that we've talked about, rash, diarrhea, stomatitis. So that's what's seen in the class. And then on the CDK4/6 side, it is on the cytopenia side. We know from CDK4/6 trials that have been conducted to date that there are very high rates of neutropenias, both in patients that are CDK naive as well as CDK experienced patients. And so we wouldn't expect to see something different than that in our trial because that's been seen in multiple trials across the board that that's a significant toxicity associated with ribo.
Great. Maybe I'm not looking for guidance here, but in terms of dose reductions or any of those things, will those be numbers that are important or could they show us some sort of differentiation in the beginning?
Yeah, so I mean, I think that, again, if you go back to the goal of what we're trying to achieve here, we're trying to demonstrate combinability. And so we'll be looking at the totality of the data and the profile and so everything that's brought to the table. The other part is that, as I mentioned, we're also focused initially on trying to define a dose. And so we've gone back, done a lot of dose escalation. And so we'll be looking, again, to define that dose at the combinability. So it's all of those pieces that we'll be looking at. I think it's hard to just kind of peg one thing per se.
Great. And then maybe with this initial data set, how can you point towards some potential differentiation from the Roche triplet here? Is there a benefit of using your ribo, which you're using, versus palbo? And is there anything that we can already see from this initial data set?
Yeah, so I think the first key differentiation with Inavolisib that we've already touched on is just the selective nature of RLY-2608. And so we don't have the hyperglycemia liability that i navolisib has. And so that allows us to go into a much broader patient population. And you can see that in the enrollment criteria and our metabolic enrollment criteria are much broader and less restrictive than the inavolisib ones. So that's one part of it. And then your question on the ribo versus palbo part, I think since ribo presented OS data, it has become the preferred treatment for physicians. And so what we hear from the clinical community is that they'd rather us combine with ribo as that's their preferred CDK4/6 today. So we just think that that physician preference is important when you're trying to enroll a study and have them participate.
But then I think the other differentiator for our triplet. We've talked about the Pfizer collaboration for atirmociclib or CDK4. And that's where the CDK4/6 space seems to be going. And so we're so far the only ones that have that collaboration and are generating data there. And that's most likely the regimen of the future is going to be a CDK4 rather than a CDK4/6.
Great. Very exciting. Well, thank you, Imogen. Thank you, Megan. We're unfortunately at time, but this was very helpful, and thanks so much for joining us.