Relay Therapeutics, Inc. (RLAY) Earnings Call Transcript & Summary

Thank you so much. My name is Peter Lawson. I'm one of the biotech analyst at Barclays. I cover SMID-cap predominantly oncology-related names and have the pleasure last few years covering Relay Therapeutics, and I've got up on stage Sanjiv Patel, CEO; and Pete Rahmer, Chief Corporate Development Officer.

And I guess the first question I've been asking a series of people have just been around the impact of the current administration. So kind of any impact from supply chain disruptions you're anticipating or worried about? And I guess the obvious parallel could be COVID, but I'd love to hear your take.

Yes. Well, first of all, thank you for the invitation. I think, as you know, I mean, the weight of news and changes is kind of unprecedented at the moment, very similar to the time back 5 years ago when we were dealing with COVID. I think it taught us a lot of lessons back then to try and diversify our supply chain and make sure that we didn't have a kind of single point of failure. And so as yet, we've not seen any real impact on the work that we do. But I think we are, as an industry, better set up having been through those COVID years to cope with anything that comes our way. But as yet, we've seen no real impact on the work we do day to day.

Got you. And then I guess the other question around the FDA and if there's layoffs or threats of layoffs, change in leadership, if that's changed the communication pattern in any way or speed or process there?

Yes. Again, as you know, like these things are very kind of discrete events. And obviously, we had a very successful end of Phase II meeting for 2608. We continue to have dialogue with them on some of our other programs. So again, we've seen absolutely no difference, but it's going to be very hard to judge and it's going to kind of be an industry-wide question. But as yet for us, no difference.

Okay. And then I guess, final kind of macro question would just be around NIH budgetary cuts, if that has any kind of near to mid-term impact on the company and also clinical trial sites.

Again, I think, again, like very, very kind of acutely, we've seen no impact. Obviously, as an industry, we don't think this is going to be positive for us and that a lot of the innovation that we built our company on has been funded at some point in the past through these grants. And so I think for our industry, this is definitely not a positive. But hopefully, this is a short-term reversal and that we can get back to something that's closer to normal.

Got you. And then I'd love to talk around -- maybe I can dovetail into this question around 2608, the trial design. But before then, maybe just like Avenous' oral SERD data, how that changes the landscape for you. They didn't hit significance on the intent-to-treat population, was it a positive, negative for you?

Yes, absolutely. I mean obviously, the VERITAC-2 trial, the readout was this morning at the top line. Those of you who didn't follow it closely, I think the market and patients were all looking for a broad kind of ITT win, which is what we did not see, unfortunately. And it was like the other oral SERD trials that have read out over the last couple of years, was positive in the ESR1 mutant population. So I think what we were hoping for is one of these SERD kind of steps forward and really truly differentiates. And this one looks like more of the same, which is they're going to be niched into later-line ESR1 mutant only patients, which is still a reasonable market, kind of 35% to 40% of those patients have ESR1 mutation. For us, I think what it means is, our choice of fulvestrant as the choice of our combination partner is one that is going to be valid for a while. There isn't going to be a SERD with a broad label anytime soon. And so the trial is kind of future-proof there. And I also think it opens up the window for us to be used in earlier lines. The worry has always been that would you be able to have a true 3 branded drugs as a triplet given the financial implications of that. We don't think that the oral SERDs will be kind of broadly used with the data that we've seen in earlier lines and therefore, opens up an opportunity for 2608.

Is there any kind of read-through Avenous had a degrader, so it was potentially dissimilar to the other oral SERDs, but in fact, it looks a lot more similar to the other oral SERDs than we thought about a week ago essentially or 6 hours ago.

Yes.

Is there any worry for that for the PI3Ks that you've got a targeted PI3K, but longer run, it starts looking more and more similar to a PI3K?

I don't think we see any read-through there. I mean I think on the true differential mechanism of action of Avenous is degrader from the other SERDs, it was a real question. Would a different mechanism of action really lead to a different outcome? And I think what we saw this morning is no. I think the question around would a mutant selective PI3K inhibitor look different from a non-selective PI3K inhibitor, I think we've answered the question. The data we showed at San Antonio showed that you get a truly differentiated safety profile, early, 9.5 months’ worth of follow-up, but still enough to show very minimal rates of hypoglycemia. And obviously, we saw a dear doctor letter yesterday night for inavolisib from Roche, showing that the problem of diabetic ketoacidosis exist in those patients. And so hypoglycemia is real in the non-selective inhibitors. So I don't think we see this as any risk in kind of the mechanism of action will prove out later to not be valid. Mutant selective inhibitors lead to truly differentiated safety profiles. And then what we saw at San Antonio was the data then translated into truly differentiated efficacy.

Perfect. That's really helpful. And kind of going back to your trial, you reduced the recommended Phase III dose...

Yes.

From 400 to 600. Just if you can walk through the rationale there and kind of pill burden, et cetera.

Yes. And so all of the work that we have done in the public domain up until a couple of weeks ago was dose fasted. And so as we went through dose escalation back in 2022 and 2023, all the work that we did was fasted. And then as we came into running the expansion cohorts, we ran those at 600 milligrams BID fasted. And obviously, we're using 100-milligram capsule for the pill burden with 6 capsules twice a day. As we came into 2024, we started thinking about dosing with Pfizer CDK4 Trilaciclib, which is dosed with food. And we also started thinking about dosing in vascular malformations where these patients will be younger adolescents. And so we started to do -- think about the food effect study and accelerating it. And so we did it, and we saw a positive food effect and showing that over the course of the year, doing a kind of bridging PK study across mono and combination therapies is the range of different oncology indications, solid tumor indications. We saw 400 milligrams BID with food had exactly the same PK profile as 600 milligrams BID fasted. And so given all the benefits of dosing with food allows us to dose better with combinations such as atirmociclib, it's probably better for patients, especially the adolescents. And if you could dose with not. And so we went to the FDA with a pretty robust data package of a large number of patients that we dosed fasted. And then this equivalent study that shows that 400 milligrams BID with food is equivalent and asked the FDA the question, we'd like to move forward with 400 milligrams BID fed, and they agreed. It does lead to a slightly differentiated safety profile as well. So, obviously, the pill burden does come down. And so some of the upper GI toxicities, nausea, vomiting, loss of appetite are significantly reduced. And so we think all around, this is good drug development. It did take us a little bit of time to do it, but we're in for the long-term. And obviously, we want to dose multiple pivotal trials over time in breast cancer and vascular malformations. So it was the right thing to do.

Got you. What are the next steps that we should be thinking about for starting the pivotal trial?

I think we're kind of all systems go now. We're obviously recruiting the team to do this clinical operations, clinical development, medical affairs, patient advocacy. We want to get this trial up and running and recruited as rapidly as possible. We'll be using a global CRO to help us all of the contracts that we need to put in place, all of the different vendors, getting the supply in place. Obviously, we're going to go head-to-head with capivasertib and so getting the CAPI supply across the world. So it's all of the kind of mundane dull things that you would need, but our entire company is focused on moving as fast as we possibly can here.

Got you. But there's no unexpected hurdles there or?

No, absolutely not.

And then the hierarchical testing, kind of just the rationale before behind the domain kinase domain and then drop down to intent to treat?

Yes. A couple of points behind the reason for that. One, we know there are sponsors that exist today with kinase-only drugs. So H1047R selective molecules, and we'll be bringing those -- they'll likely bring those forward into the clinic and potentially can end up sometime in the future with labels for kinase-only patient populations. And so therefore, by having the kinase hierarchical testing, it allows us to potentially, upon a successful study to be able to make direct label claims on the kinase population. Additionally, we have seen obviously some activity -- the kinase population has been performing even better than the total population, which is already almost doubling the median PFS of the control arm. And so therefore, we are -- we built this study to make sure that we win on multiple different parameters. And so we built it to have a bit conservatively on the end, and we built it a bit conservatively on having this hierarchical testing, knowing that in the worst-case scenario, if we just had the kinase population, that'd still be a huge market for us, too. But we have -- the base case is complete confidence in the total population and then the ability to have a -- to make some label claims to the kinase population on the other side of the successful study, knowing that there will be sponsors with kinase-only drugs coming behind us.

And what drives that ORR difference that we saw between the mutant domain or patients with mutant domain and then patients with the non-kinase mutations?

Yes. I think, look, there's -- when you get into sub-populations analysis of a study this size, you get into some imperfections in looking at different patient populations. So for instance, in the non-kinase population, those patients in this study tend to be a bit more heavily pretreated. So they're more likely to have seen multiple CDK4/6 inhibitors. They're more likely to have seen fulvestrant and multiple endocrine therapies. They're more likely to have seen chemotherapy. So, I don't know as though we can say for certain that there really is any performance difference between the kinase and the non-kinase population because when you then look at the PFS, our second-line 9 kinase patients perform almost exactly the same as our total population, median PFS of 9.2 months. So I think we really need to get into a larger study like we're about to do and probably that will be the place where we can start to really learn in a well-controlled way if there's any difference in the performance of these patients. But on the surface, we don't think there's going to be a meaningful difference between the 2.

But in terms of just the kind of reference numbers there, obviously capivasertib is doing 5.5 months of PFS in that population. We were saying that the overall and the non-kinase are in the zone of 9.2. So there's still a significant buffer between this kinase is doing even better than that. So in the kind of head-to-head randomized controlled trial, we feel very confident that both sub populations will do significantly better than the standard of care.

Got you. So we should more so probably think about it as disease control rates because that probably translates better to the PFS number.

Yes. Clinical benefit rate is definitely a much better predictor of eventual progression-free survival. And I think in the case of our study, the ReDiscover 1 study, what we are seeing is a bit more consistency between kinase and non-kinase when you look at CBR and PFS. Yes, the response rates do differ as it stands today.

Do you think that also the similarity in PFS, does that also translate to OS as well? Or do you think there's a delta there when you think about the kinase versus non-kinase?

As you know, OS measurements have other factors that come into play, the subsequent treatments, the geographies you exist in, maintaining balance and making sure that you have similar follow-on treatments that these patients go on to. I think on an absolute basis, the -- assuming all else being equal as to what treatments these patients go on to afterwards, we would also expect not too much of a difference between OS performance between the 2 populations.

How should we think about as the trial enrolls, time lines for enrolling and kind of running that trial?

Yes. I think what we can say today, not having enrolled a single patient yet is if you look at some precedented studies in this space that sit in this 500 patients end, they take about 3 years to get from first patient enrolled to top line data. And I think we might be able to refine that as we get a year or 2 into the study and really understand pace of enrollment and -- but I think as we sit here today, pointing that precedent is probably the best we can do for now. Obviously, as Sanjiv said, this is the sole focus of our company. And so we'll use every tool at our disposal to accelerate this as much as possible.

Got you. And then the pretreatments the patient or prior treatments the patients have had, would you include patients with prior CAPI, other PI3Ks, kind of just how should we think of how that shakes out?

Yes. There -- we don't allow prior treatment with pathway inhibitors onto the study, similar to the ReDiscover study. We do allow up to -- we require 1 CDK4/6, at least 1 CDK4/6. That can be either in the adjuvant setting or in the metastatic setting, up to 1 or 2 prior endocrine therapies. They can see up to 1 -- patients can see up to 1 round of chemotherapy in the metastatic setting, and that's kind of to allow this ability in certain geographies to -- in the metastatic setting, if a patient is coming on to the study, but there's any mechanics waiting for a diagnostic test to come back, you can allow for a couple of cycles of chemotherapy to -- and those patients would still be eligible to be into the study. So there's -- I think what we're really trying to solve for is getting to a patient population that looks and feels like a true kind of second-line metastatic patient population as where in the ReDiscover study, as you'll recall, we had about 45% of our patients were third line plus.

Yes. Okay. Perfect. And then can we get the full Phase I data some point this year, I guess, I assume like towards the year-end. Is that -- how should we think, one, the timing of that data? And then what should we be looking at in that data set?

Yes. They go hand-in-hand. As you recall, the last update we gave was at SABCS, so just in December 2024. And that was from a data cutoff of early November 2024. So we want to make sure whatever next data update we give is enough time to let that data mature a bit further. The median follow-up at that time was 9.5 months. And so we think letting the data get to in and around about a year's worth of median follow-up is probably a meaningful update to the existing data. And showing consistency of within the median PFS and even further maturity probably would give us all comfort in continuing to increase the POS of the Phase III study. So I think that's how we think about that update. Still not answering your question directly on timing, but you can kind of hear some of the thought process that goes into what we're hoping for by the time we show that data.

Is that something you would have submitted to ASCO or?

We don't really talk about our abstract submission strategies. I think ASCO and SABCS are rational bookends for where this update would make sense. As you can imagine, one of the key things -- one of the key reasons for doing this update is in support of awareness for the Phase III study. And so putting it at a major global medical meeting is -- would be prudent to continue to raise awareness and excitement around the Phase III.

Okay. Perfect. And then just as we think about your data versus Scorpion and kind of where do you think you are? Who's ahead by how many months or years? Or how do you think about that?

I mean I think all we've seen so far from Scorpion is kind of the dose escalation monotherapy data sets that they have. And so those are obviously done in a range of different tumor types, and I think a follow-up over close to 2 months. So very early. Obviously, very kind of similar to our data set that we showed back in early 2023 at AACR, where we had roughly 30 or 40 patients, but we did out in mainly in combination. And so if you think about what they need to do and obviously, dosing combination, get very kind of good, robust long-range follow-up against different expansion cohorts before they kind of set to go and do a pivotal trial. Obviously, they've been through a sale process and an integration, which is probably still ongoing. So we think we're somewhere kind of close to 18 to 24 months ahead. And I don't think we're going to see anytime soon a robust, long-term data set in combination with fulvestrant in hormone receptor positive HER2-negative breast cancer patients that you could compare with what we've got.

Yes. I think to be very clear, that molecule is at best similar and it's 18 to 24 months behind. There's been a lot of conversations around target coverage, safety, efficacy. And if you go right down the list, you can't find any credible set of data that supports any meaningful differentiation. And to the contrary, many open questions that remain for that molecule, ones that we've already answered pretty clearly with our longer-term follow-up mature data.

And as we think about your triplet data, have you seen any kind of drug-drug interactions between 2608 and rivo?

Yes. We've discussed the rivo1 last year. We are a victim of rivo, where rivo is increasing the exposure of RLY-2608. And so therefore, to get to the equivalent exposure of the 400-milligram BID fed likely requires a lower dose in combination with rivo. We're still working through that. We know that, that DDI exist. That study was started in the early part of 2024. We first discovered that once we discovered the DDI with rivo, we had to kind of take our time and lower the dose and understand that. And then we are moving forward all our -- all studies moving forward will be in the fed state. So that's also changing -- shifting that over to being in combination at doses in the fed state is also something that we're doing. Obviously, with -- as Sanjiv alluded to, dosing trilaciclib fed is important given that, that drug is also dosed fed.

Does that confound dosing in the real world in any way, just having that interaction and if patients drop off of one of the combination agents, does that deturb or change the way physicians think about using the drug or dosing the drug?

Specifically with ribociclib, it does bring that into consideration because obviously, if for some reason, if a patient went off rivo for a week, it's going to change the exposure of 2608, but it's something we think through and have been working through. The benefit of trilaciclib is that's dosed continuously. And obviously, they're now both being dosed fed and it's you don't have some of the safety and side effect issues that you have with rivo. So it's an exciting combination to be exploring.

Remind us where you are for dosing and when you start generating the recommended Phase II data?

Atirmo was just more recently started at the very, very end of last year. So this year is all about execution of those combinations, and I think sets us up well to probably be able to talk more about these triplets as we get into 2026.

Are there other combination agents we should be thinking about? I guess there's potentially one less today.

Yes. I mean if you go full circle back to this conversation, we imagine the oral SERDs will be present in kind of later lines in the SR1 mutant patients. And so you'll see us do some combination with these oral SERDs so that we can be present in some of these indications. Obviously, the choice of those is going through our thought process at the moment as we see the data.

Got you. We've got 30 seconds on the clock, vascular malformations. Just kind of where are you in that process? What type of patients will you initially enroll in the particular subsets?

It's a really exciting opportunity for us, over 170,000 PIK3CA mutated vascular malformation patients in the U.S. alone. The initial study is set to start imminently. It will -- the alpelisib's current accelerated approval is in a specific subset, PIK3CA related overgrowth syndrome. So we'll have probably some bias in our initial study, initial randomized dose exploration towards those patients. They also tend to be a bit more of the severe patients. So, ones kind of seeking systemic therapies, especially early experimental ones in a first-in-human study. But we see a broad opportunity across all the different sub-types of vascular malformation, pros, lymphatic malformations, venous malformations and cerebral cavernous malformations over time.

Perfect. Thank you so much.

Thanks again for the invite.

It was a pleasure to speak.