All right. Welcome back to the Citizens JMP Life Sciences Conference. My name is Silvan Türkcan. I cover precision medicines. Welcome back, and thanks for joining us today. I'm here with Revolution Medicine . I'm with Imogen Pyrce , COO of R&D. [uncertain] , thanks so much for joining us today.
Thanks for having us.
For those not familiar with your approach to drug design and how you pick opportunities, could you just walk us through how that works?
Yeah, absolutely. So Relay Therapeutics was founded in 2016. We sit at the intersection of leading-edge experimental and computational tools to create new precision medicines and ultimately to transform drug discovery. Our approach and Dynamo platform looks to break a drug discovery problem down into tractable subproblems that are amenable to solution with our platform and ultimately look at AI sitting on top of an experimental foundation to solve the problem and create value for stakeholders. Our platform drives the approach, and so we are looking at the platform itself melds different computational and experimental tools together in order to go out and solve different drug discovery problems across the continuum. We've validated our approach. So over the past eight years, we've put four precision medicines into the clinic.
Our strategy for choosing targets is based on four main factors, the first being to look at well-validated protein drivers of disease. The second is to look for places where patient populations are clearly identifiable. The third is low translational risk targets. And then finally, we look for places where we can have a rapid path to clinical proof of concept. We go after problems that are hard, where others have tried but not been successful in solving them, so-called intractable problems. And we look to minimize biology unknowns as much as we possibly can. Our approach is differentiated in how we integrate experimental and computational tools and techniques, and most importantly, by our team and how they work together.
Great. And so your primary focus at the moment is RLY-2608 PI3K alpha inhibitor, the mutant-selective. Could you just walk us through the rationale behind the mutant-selective targeting and what evidence we have that that's a good idea?
Yeah, absolutely. So the existing agents are non-mutant selective, which means that they inhibit both the mutant PI3K Alpha, which is responsible for tumor growth, and also wild-type PI3K Alpha, which is responsible for a lot of normal physiology, including glucose regulation, glucose metabolism. So the challenge with these existing agents is that because they are equipotent against both the mutant and the wild-type form of the protein, they suffer from off-target toxicity. In this case, targeting wild-type PI3K leads to hyperglycemia, which makes it a challenging safety and tolerability profile for patients, makes it much harder for patients to stay on drug, and ultimately limits the efficacy of these agents.
Great. Over the years, you've generated a lot of preclinical data with this compound. What are kind of the highlights that investors should know about that really demonstrate the fact that hyperglycemia is not a problem with your asset and some of these platelet-related toxicities are mitigated?
Well, so as you mentioned, we demonstrated preclinically that that should be true, that we should have a better safety and tolerability profile. And then we showed that this translates into the clinic. So in our disclosure, in our most recent disclosure last August, we were able to show that we had no G rade 3 or higher rates of hyperglycemia in patients. And this then translated into patients staying on drug. So at the time of the data disclosure, 15 of 17 patients remained on the study, and 6 of 7 CBR evaluable patients had achieved stable disease or better at 24 weeks.
Great. Yeah, maybe jumping right into that data. So you had an update at AACR last year, and then you updated that at the summer. Obviously, in the beginning, there were maybe perhaps fewer responses. And then in the summer, you started seeing some. So could you just maybe give us the high-level takeaways from that data and how it evolved from AACR to the summer?
Yeah, at AACR, we showed that we had multiple therapeutically active doses, that we were able to dial out the off-target toxicity that Imogen just talked about, in particular hyperglycemia, rash, diarrhea, which are the three key AEs that cause patients to discontinue treatments. So we showed that profile at AACR. Then later in the summer and August, as you said, we disclosed that we had selected the 600mg BID dose to bring forward into the first dose expansion cohort. That was based on the data that we had seen from the 17 patients at the 600 mg dose and the dose escalation. So at the time, as Imogen mentioned, 15 of 17 were still on drug. 6 of the 7 CBR evaluable patients had stable disease or better at 24 weeks.
4 of the 5 patients with measurable disease that were efficacy evaluable had tumor regressions as well, as well as one confirmed partial response. It was really the combination of that safety and tolerability profile as well as the early efficacy data that gave us the confidence to take that 600 mg dose forward.
What stuck out to me in the August update was that the duration of treatment, so not necessarily response, but treatment, was quite long. Did you know offhand what that number was?
I believe the median duration was about in the 12-week time frame at that point. But we had some patients that had enrolled fairly recently, but then we also had patients out much longer too, kind of the nature of a dose escalation.
Since then, what was the progress with the ReDiscover trial and what cohorts are now enrolling?
So right now, so last fall, we finished the dose escalation portion of the combination study with fulvestrant. We then chose 600 mg BID . as a dose to move forward into expansion. We opened an expansion cohort at 600, completed enrollment there, and opened another dose expansion cohort at 600. We've also opened a 400 mg expansion cohort, largely due to Project Optimus reasons, to try and fulfill that requirement.
Great. And then the trial now includes also the triplet, the 2608, fulvestrant, and then a CDK4/6 arm. What are the plans for moving the triplet into early lines, perhaps? At what stage will that happen?
Yeah, so one of the things that's most exciting about the safety and tolerability profile that we've shown with 2608 to date is its combinability with other agents, such as CDK4/6, and thus the potential to move into earlier lines of therapy. And so to investigate that, we've started triplet dosing with fulvestrant and ribociclib earlier this year. To date, none of the pathway agents have been able to combine with ribo. And so that would be differentiating if we can show that. And so we're doing some early dose escalation there. And we'll expect to show data from that early dose escalation in the second half of the year. And that readout will focus on, as I said, it's dose escalation. So it's small numbers. It's going to be short duration. And patients are going to be largely post-CDK4/6.
It'll really be focused on giving us an initial read on safety and tolerability and ultimately combinability with a CDK4/6.
Just one important thing to note there is that we will be dosing patients who've already seen a CDK4/6 inhibitor rather than some of the intervals of study that we saw readout at SABCS was in pristine CDK4/6 patients. So that's something that we'll have to keep an eye on as we start to think about that data.
Great. There was a slight deviation from the, I guess, what you would call the natural distribution as the patients enrolled that had kinase mutation helical patients versus other patients. Is that important? And did you mitigate that in any way? Or do you think it's just more interesting to have a different?
Yeah, so for context for the audience, in the natural history, about 90% of patients have what are called hotspot or helical or kinase mutations. Then about 10% of patients have so-called other mutations. And so as the only study that was open to helical kinase as well as other mutation patients, we saw a bit of an enrichment in enrollment from patients with other mutations. And so in the second of the 600 mg dose escalation cohorts that we opened, the one that we opened this year, we are restricting that to helical and kinase-only patients in an effort to get back closer to that 90/10 that we see in the natural history of the disease, and additionally to make sure that we have a data set that's comparable with the existing agents, which were studied really just in helical and kinase patients.
We want to, to the extent that you can do cross-trial comparisons, obviously, but a data set that's more comparable to that.
Great. And obviously, you've already mentioned this, but in the second half of this year, we'll get a fairly sizable update from the doublet and the triplet. Could you just maybe outline, I don't know what the best way is, but maybe patient numbers or what doses and how much follow-up, and how meaningful will this data be in determining what's the recommended phase 2 dose and then what comes next?
Yeah, so starting with the doublet, as we've talked about, we demonstrated differentiated safety and tolerability last year. And so what we're really focused on in the second half data readout for this year is the more duration on the efficacy side and a more robust read there. And so in terms of patient numbers there, we'll have the 17 patients from the 600 milligram dose escalation portion as well as the 20 patients from the first dose expansion cohort at 600 mg. So taken together, you have nearly 40 patients that will have at least six months' worth of data. And then you have the second 600 milligram cohort of 20 patients that's currently enrolling. We'll report those data as well. But obviously, the duration will range a little bit more from there.
They won't have had quite as much time on drug as the 17 and the first 20. But taken together, as I said, you'll have the nearly 40 patients that are CBR evaluable. So that'll really be where the focus of the data are. But then we'll also be looking. It's unlikely we'll have a mature median PFS by that point. But we'll endeavor to do some landmark analyses at six months, potentially at nine months, to give us a directional sense of kind of how that's trending as well.
Great. And you mentioned that the second cohort, was that the 400-milligram cohort or the second 600?
The second 600, yeah. So our guidance for the second half doublet data update is specific to the 600 milligram cohort. Yeah, we haven't guided to the 400 yet.
OK, great. That's very helpful. What is a good outcome for efficacy and safety in that doublet cohort?
Yeah, so the regulatory endpoint is median PFS. And so there's two approved agents in the space, capivasertib and alpelisib. And with those agents in the post-CDK4/6 patients, so in that true second-line patient population, the median PFS is between 5.5 and 6 months. And so that's what we see as the bar to beat, is to have data that show that in a pivotal study run against one of them, we could comfortably beat that 5.5-6 months, which is the other reason that the clinical benefit rate, which is a measure at 24 weeks or essentially six months, is so meaningful as well.
And what's the bar for discontinuations? Obviously, some of these safety profiles you've already said, you're already maybe superior to the other PI3K Alpha here. But is there a hard bar for you? Or is it just lower than alpelisib?
Yeah, so alpelisib is about 25% in their phase III study. And then capivasertib in the CAPItello-291, it was only a 13% discontinuation rate. But you also have to remember the context, which is that capivasertib is dosed four days on, three days off. So patients are essentially getting a weekly drug holiday of about three days.
Yeah. And maybe with respect to capivasertib versus alpelisib, capivasertib is rapidly gaining market share, I think, in the U.S. And doctors just seem to like it. Is it because alpelisib is just not all that? Or how do you view that? Or is the AKT inhibition something that just attracts doctors, the different MOA here?
Yeah, as you said, we've been hearing the same thing, that physicians are really preferring capi. And the recently reported Q1 results really show that that is playing out commercially as well. Capi or AstraZeneca reported very strong first quarter sales, about $50 million for capi. And then you also take that coupled with the quarter versus prior quarter sales of alpelisib, which were down more than $20 million. And so it shows that what we're hearing from the physician community is playing out. In terms of the reasons for it, I think it's less about the mechanism from an AKT versus PI3K perspective and more because of the fact that the capi dosing that we were talking about, the four days on, three days off, allows them to minimize grade 3 and greater hyperglycemia in particular, which is a challenge for these physicians to manage.
It's not something that these oncologists are used to managing so much as they're trading it off for much higher diarrhea. But that's easier for them to manage. Whereas once you get into the grade 3 plus hyperglycemia, you're generally having to call in an endocrine consult. And it just makes the management much harder. And so that's the trade-off. And that's what we're hearing from the physician community, is that it's that manageability of the off-target toxicity trade-off.
And so does that put you in a situation where you have to basically lower diarrhea as much as you can to compete with capi? And then, I guess, grade 3 hyperglycemia to differentiate, say, from alpelisib, is that like you would just want them as low as possible? It is not that you want hyperglycemia being lower than alpelisib and diarrhea being lower than capi?
Yeah, I think it's looking at the whole profile in context. Because while physicians prefer the lower hyperglycemia, frankly, patients don't love the higher diarrhea, as you might imagine. And so it's really a focus on the grade 3 and higher levels as well, in particular hyperglycemia. Because, as I said, that requires an endocrine consult, is much harder to manage. And so that will be the focus for us, is kind of maintaining the low levels of all three of those AEs that we've talked about, but in particular continuing to minimize grade 3 and higher hyperglycemia. Because grade 1, grade 2 is something that those oncologists are more comfortable managing. But once you get into grade 3 and higher is when it's harder for them to manage.
That makes sense. Maybe what's the bar for safety for the triplet? Is it kind of comparing it to Roche's triplet data? Or what should we look for?
Yeah, I wouldn't say it's comparing it to Roche's right now. And I'll come back to that. But I think the bar of what we are looking to see is continuing to see the low rates of grade 3 and higher AEs that we've been talking about, and in particular looking that there's no additive toxicities, meaning that the profiles of the individual agents are really just coming together, that you're not increasing rates of any of the key AEs you see with any of the individual agents. And then in terms of the question about comparability to Roche, I think there's a couple important pieces of context to look at there. The first is just which lines of therapy the studies are being conducted in. And so the Roche study with inavolisib, the data that they reported at SABCS, was in a true frontline patient population.
So they couldn't have seen a CDK4/6. Whereas our patients have largely all seen CDK4/6. So you have CDK4/6 refractory patients. They've seen or they've been able to see up to at least up to one line of chemotherapy in the metastatic setting and then also have seen prior endocrine therapy. So just very different levels of pretreatment in those patient populations. And then the other piece of it that would make comparison hard and not necessarily appropriate is just that the Roche study very strictly controlled the metabolic status of patients coming into the study. And so our study is allowing a much broader patient population. They excluded all prediabetic as well as diabetic patients, which the CDC estimates is about half of US adults. Whereas we're allowing a much broader patient population into the study.
So I wouldn't say that we're looking to compare across those studies, just given all those different caveats of different patient populations.
Great.
Yeah, I think the only other thing to mention on this front is that our initial triplet study is with ribociclib, which other existing agents have not been able to combine with because of the overlapping toxicity. So it's not even like, what is the toxicity? They just haven't been able to do it to date. So it'll be interesting to see how that data plays out.
Yeah, so if that safety data looks good, it could be helpful in designing a frontline study. Because you could potentially design something with combining it with any CDK4/6 rather than specifying one. Or is that something you'd want to do?
That would be the goal, I think, would be to be able to use any of the CDK4/6s that are out there.
Our current protocol allows for palbo, too. That's something you could see us add in the future as well, just to make sure we're looking at both. The Roche study is combining with palbo, just as a reminder for everyone. Novartis did try and combine alpelisib and ribo previously, but they couldn't because of toxicity.
Great. I want to ask you about another asset, LOXO-783, that only targets H1047R mutants. What is your thinking around that asset progressing? I don't think we've seen data today. But if there is data.
It's a good question. It went into the clinic only a couple of months after 2608. So it's been in the clinic more than 2 years now. But as you said, we haven't seen any data. It targets H1047R specifically, which is about 40% of patients. So our panmutant approach allows for a much broader patient population. What we do know preclinically is that it's very selective. What we don't know is whether that translates to any differentiated efficacy in the clinic. And until we actually see clinical data, it's hard to speculate on that part of it.
Great. Maybe thinking about after we see some of this data and for the doublet now, how do we get to a registration-directed trial for the doublet? When is that more finalized?
Well, I think as we've been talking about, we're going to continue to see the data mature. In the back half of this year, we'll look to go to the FDA and align on a pivotal path forward. The earliest, as we've been discussing, the earliest pivotal trial we could envision running would be a second-line trial. We'll have to get agreement with them on both dose and comparator that makes sense for that study.
Great. And we have about a minute and a half left. So I want to also highlight the fact that you have several other great assets. Obviously, a SHP2 inhibitor, which we may or may not be able to talk about. But maybe, first of all, lirafugratinib, which I might have mispronounced. But it's FGFR2-directed asset. And you had really good data in cholangiocarcinoma. And obviously, then you expanded in several others. And now this is enrolling in the background. Could you just describe maybe the decision that led to you just keeping on enrolling it? And what should we think about that asset's future? Is it important? Is it not important? Will you bring it forward or not?
Yeah, so as you said, we showed some promising tumor agnostic data at the Triple Meeting last fall. They were early, though. And so what we've said is that we just need more data and more maturity. And so after the Triple Meeting, we continue to enroll patients. We now feel like we have a sufficient number of patients enrolled. So enrollment has been closed. But we are just letting those data mature. We've also said that we've minimized spend while that's happening. And so we'll let those data continue to mature. And then we will take them to regulators, discuss them, and potential path forward. And so what we've got it to is the second-half update for tumor agnostic regulatory update, as well as data. And to your question on kind of our potential plans, I think it's shown to be a very clinically differentiated asset.
But we also know that tumor agnostic commercialization is very hard. Even clinically differentiated assets have had a struggle commercially. And so for us, the bar just continues to be very high in terms of the money that we're going to be able to put against it. It's a great asset. But we also have a lot of other great assets, the PI3K program, obviously at the front, very large opportunity. But we'll also be disclosing some currently unnamed programs on June 6, as we just announced yesterday. And so we're very excited about some of those new programs. So it's just kind of making sure we're prioritizing spend across the portfolio as well.
Great. Well, thank you so much. With that, we run up on time here. But thanks so much for joining us today.
Thanks for having us.
Thanks.