Hello again, everyone. My name is Etzer Darout. One of the senior biotech analysts at Barclays. It's my pleasure to have our next Fireside Chat with Relay Therapeutics. With us this morning, we have Sanjiv Patel, CEO of Relay, and Peter Rahmer, Chief Corporate Development Officer. Thank you so much for joining us this morning. Maybe, Sanjiv, just to get us started, maybe a brief overview of Relay Therapeutics for those maybe less familiar with the story. Yes. Well, first of all, thank you, Etzer, for the invite, and thank you to Barclays. It's always a pleasure to be here. The company's been around now close to 10 years. We're a company that's focused on trying to drug precision medicine targets. We have a very exciting year ahead. It's been set up because over the last couple of years, like every company in our industry, we've weathered you know the very tough environment biotechs have been through. We weathered it in our own way by taking our lead program, which is an FGFR2 inhibitor, and out-licensing that. We very much focused our research organization down. We staggered some of our preclinical programs, and we focused the majority of our resources on our PI3K alpha inhibitor, which we believe is the first mutant selective inhibitor to enter the clinic, zovegalisib, on three big opportunities. Those three opportunities are hormone receptor-positive, HER2 negative front line and second line metastatic breast cancer, and then PI3K alpha-driven vascular malformations. All three are we believe to be very large commercial opportunities. Over the coming year, we're going to show three datasets. First of those datasets is in second line metastatic breast cancer, which next week at ESMO TAT, we'll show data from our pivotal dose, 400 milligrams BID fed. All the data we've shown to date in the public domain has come from the 600 milligrams BID fasted cohorts. This will be the first time we'll show this data. It's 57 patients worth of data. There should be enough follow-up to assess both response rate and PFS. The goal is to show consistency. We've in the past shown 10 months worth of PFS in this patient population, where the standard of care is at 5.5 months of capivasertib. If we can go some way towards showing consistency, that should give investors great comfort that we will be able to run a very successful Phase III trial that's ongoing, the ReDiscover-2 trial. The next area is frontline metastatic breast cancer, where in 2026, we'll show data around which frontline regimen we want to take forward. Currently, we're testing three different regimens using both palbociclib, ribociclib, and Pfizer's selective CDK4 atirmociclib. We will declare which of these regimens we want to take forward. Our prioritization is going to be made on tolerability. We're looking to get the most tolerable regimen we can. The efficacy, we believe, has been shown by previous trials like Roche's INAVO120, that a PI3Kα inhibitor in the front line is going to add efficacy. They've obviously shown overall survival benefit. The challenge really is getting a triplet regimen that shows tolerability. We're hopeful sharing which regimen to go forward with and what that trial design is. Then the third area is vascular anomalies. This is an area that's very poorly kind of understood by most investors, but it's a very large commercial opportunity that's unserved. It has a few approved therapies, most notably alpelisib from Novartis has an accelerated approval in a sliver of this patient population PROS. What we would like to show is in the first half of this year, 20 patients worth of efficacy data. That efficacy in this population is measured through volumetric MRI reduction of lesions. To get a response, you have to have a 20% reduction in lesion volume. The bar has been set by Novartis in their accelerated approval trial at 27% response rate, and in their confirmatory trial was at 17%. We'll show data at a slightly earlier time point, 12 weeks in these 20 patients, and we hope that we can show proof of concept here and then move rapidly to next steps. We have about $550 million of cash at the end of last year, so we're well set up, over the coming years to generate value for shareholders, in these three very large commercial areas. Great. Thank you for that. Maybe just for our listeners to just maybe kind of put into context the role of mutant PI3 kinase in breast cancer. Obviously, it's a space that's rapidly evolving and where you sort of see the value creation for having a mutant PI3 kinase selective PI3 kinase in these populations. As the hormone receptor-positive, HER2-negative breast cancer market is very large. 40% of those patients have a PI3Kα mutation, and it's a foundational mutation. It's truncal, it's present at the diagnosis. It's not something that emerges through therapy like ESR1 mutations. We've always believed that the precision oncology has shown us as well that if you just target the driver, which in this case we believe the driver is PI3Kα, you should be able to see efficacy. We've seen that time and again in successful medicines in our industry. For decades now, our industry has tried to make a mutant selective inhibitor. The challenge of the previous generations is that they either hit the isoforms of PI3K, alpha, gamma, delta, and just alpha, or they hit the wild type as well as the mutant. This leads to a range of off-target toxicities. Most notably, hyperglycemia, diarrhea, rash, stomatitis. Although the efficacy is clear, unfortunately, the tolerability of these prior generations has been challenging. Patients just can't stay on therapy, and that leads to reduced dose intensity, and that unfortunately leads to poorer efficacy. Mm-hmm. The goal of a mutant selective inhibitor is can you hit the target hard, make sure that you get the efficacy that patients need, but remove some of the off-target toxicities to make it a tolerable regimen. We believe that's exactly what we have achieved with zovegalisib. The data we've shown to date shows that we have very low grade 3 hyperglycemia, diarrhea, rash, dermatitis. We've kind of moved away from some of the previous generations' challenges. That, in the second line, has shown a PFS in the data that we've shown of almost two times what we see in the standard of care. We believe that will also translate into a frontline regimen that should be tolerable because you're removing some of the toxicity. Obviously in the vascular anomalies as well, these patients are going to be chronically on therapy. Again, the vascular anomalies is driven by PI3Kα mutation, so if you can cover the mutation above the IC80 or even close to the IC90 for 24 hours, you should be able to see the efficacy. Given these patients are going to be chronically treated for many, many years, the tolerability is paramount. That's why the industry has always wanted a mutant selective inhibitor, and I think we are the first to deliver it. Right. I think the data abstracts I believe will probably be published tonight for your upcoming Yeah. Presentation. What can we expect, I guess? Would we get more data at the presentation relative to what we see at the abstract? Maybe if you can also remind folks what we can expect to see during the presentation. Yeah. You're correct. The abstracts are slated to come out tonight around 6:00 P.M. or so. Mm-hmm. The abstract will contain the top-line data from an earlier time point. Mm-hmm. It'll have PFS response rates and the PK data. 'Cause recall what we're largely doing here is trying to show this translation of the 600 fasted dose to the 400 fed dose. Mm-hmm. It'll contain all of that. Right. The data presentation itself on Monday will have a later time cut, but a lot of the same headline numbers. Mm-hmm. I think you'll see the key takeaways. Again, the goal of these data is to show consistency from the 600 fasted data. Right. Great. obviously we field a lot of questions around comparisons with Cellgevity's molecule. Maybe just if you're thinking about sort of the differences between those molecules, those programs, and then also the data that they presented to date relative to what they could show at that mutant population, which people are anticipating greatly from them. Yeah. I mean, as we talked about, we have a very kind of targeted therapy which hits PI3Kα, and you know has selectivity between the mutant and the wild type and essentially leaves everything else alone. As far as the biology of the selectivity, get it totally. But you know it has broader coverage and you know hits multiple nodes in the pathway. They have obviously shown you know great data in the wild type and obviously moving very fast towards an approval there. That's great for the field because in the wild type, therapies just aren't providing the efficacy that's needed, and this will be a step change for patients. I think in the mutant data that's upcoming, the real challenge is just going to be it is an IV therapy weekly. It does have a range of toxicities that do come from having multiple nodes hit in the pathway. The challenge is just going to be how good does that data have to be above both the standard of care today, and so 5.5 months is the capivasertib bar. Mm-hmm. beyond an oral therapy that could emerge in the field like us in and around the 10-month PFS. How high does the efficacy have to be above that to justify the IV and the toxicity that goes with this. I think obviously the commercial markets will provide the answer to that over time. I think what we can do is provide a clean as we can tolerable therapy with as high a PFS as we can that's convenient to patients that has an oral route of administration. We hope that all the work that we've done with both physicians and patient groups that's a very attractive profile to have. Right. In terms of the enrollment of your Phase III ReDiscover-2 trial, maybe any commentary around how enrollment is progressing and perhaps is the trigger for sort of the data timing would be maybe a completion of enrollment. How are you kinda thinking about when you disclose sort of the data timing, for the- Yeah. Maybe I'll take the first part and then maybe Pete can take the bit about guidance. We're very happy with how the trial is going. Obviously we started it late last summer. We're setting up sites across the world. We're enrolling patients now in every region in the world. There's great enthusiasm given both the strength of our data and the trial design itself. Obviously, we're using an active comparator in capivasertib, which given the strength of the launch of capivasertib is a very attractive proposition. Mm-hmm. Especially in countries where capivasertib is not reimbursed. given the kind of various trials that are out there, most of the trials in this field in the mutant population in the second line have kind of completed their enrollment, so we're not kinda up against headwinds in terms of other trials. Mm-hmm. We feel very positive about how things are going in terms of guidance. Maybe I'll hand over to Pete. Yeah. I think once we get all sites up and going globally and have a good sense of the steady state enrollment curve and pace of enrollment, we'll endeavor to try to give maybe some guidance to when we could reach full enrollment or a good range of when that might come. These are event-driven studies. PFS is the endpoint the control arm does in around 5-6 months. Mm-hmm. Typically in precedented studies, the top line data comes not too far after that full enrollment period. Right. I think that's kind of the kinetics of how we'll try to keep folks updated on the study. Great. No, thank you. For the front line, maybe the types of data disclosures we can expect to ultimately help decide, right, which is the right CDK4 selective CDK4/6. How are you thinking about that and what that data set could ultimately look like that help drive that decision? Yeah, obviously as we talked about, we are doing three different triplet ongoing work. The majority of that work is happening in second-line patients because that's the patients that we could most rapidly access. Obviously, we're looking for the tolerability profile above all else. We know that the efficacy of the triplet will play through. The disclosure that we'll make at some point in 2026 will probably focus on the go-forward regimen, and it will show obviously the demographics of the patients that we have. These are going to be heavily pretreated patients with multiple lines of therapy, the tolerability profile, and whatever follow-up we have at the time. It's probably going to be too early to assess PFS in these patients. You know, our thesis is very clear, which is that that side of it has already been proven. We're trying to figure out now, which is the most tolerable regimen that patients could take 3-5 years' worth of therapy. Obviously we will share the next steps in terms of the go-forward trial design that that we hope to run in the front line. There's a lot to try and communicate here to investors when that disclosure is made. Obviously, we need to show all the kind of precedent, comparative data that we're going to try and make the decision against, have investors kind of understand how we came to the decision, and then obviously share the next step. It will be a fulsome disclosure with lots. We'll probably have several bites at it over time. Yeah. That'll be the first step is to disclose those three key pieces and give a sense of timing, and then we can come back at it as we get closer to the launch of that actual study. That data we're generating in the Phase I, two setting with the triplets, that'll mature over the next few years as we get the frontline trial up and running. That'll be able to provide folks continued insights into the profile of the triplet regimen we choose. Right. One thing I haven't taken a look at is this sort of PI3 kinase expression levels prior to sort of a frontline, so like an adjuvant setting, for example. I mean, obviously would be something much longer term, but is there an opportunity to have combinations in the adjuvant or early line settings of breast cancer? Yeah, absolutely. As Sanjiv mentioned upfront, these are truncal mutations foundational to the disease. Okay. At onset in the adjuvant setting, you're going to have 40% of the patients with a PIK3CA mutation. We do think it's a driver event in those patients. as you take a step back and look at the big picture of what the opportunity is for zovegalisib in breast cancer certainly we have near-term plans for the ongoing second line metastatic disease in HR-positive, HER2-negative. We have frontline plans. You can absolutely contemplate adjuvant opportunities in HR-positive, HER2-negative disease. I think over time, we'll think about even moving into other breast cancer settings such as HER2-positive disease or even triple-negative disease. There's a high frequency of these PIK3CA mutations who require probably different combinations with different agents. Mm-hmm. Certainly an opportunity to help address those patients too. Right. Maybe switching to vascular malformations, and you talked about it briefly earlier. Just curious, where are the majority of these patients treated in the community setting? Do they need to travel to specialized treatment centers? Yeah. I would say just to start at the broader picture with vascular anomalies, there's about 170,000 of these patients. Mm-hmm. With PIK3CA-driven disease in the United States from a prevalence standpoint. The vast majority of these patients exist on a more mild spectrum of the disease. They probably are seen and being treated within their primary care physician or dermatologist because in the more mild presentation of the disease, it's more of a cosmetic cutaneous presentation that doesn't require more aggressive intervention for the disease. Right. I think what the context of the disease that we're thinking about is the more moderate to severe. There you have there's a subset where alpelisib has accelerated approval right now. It's called PIK3CA-related overgrowth spectrum. It is a syndromic form of the disease. It involves usually multiple limbs and/or tissues or critical organs. Mm-hmm. That's where you'll see the bulk of our initial data is in those patients. Those patients tend to find their way to what are called vascular anomaly centers. There's probably 2-3 dozen of them in the United States, another 2-3 dozen in Europe, where the vast majority, or at least our understanding today, is vast majority of the moderate to severe patients make their way to those types of centers. They tend to overlap with children's hospitals. Mm-hmm. 'Cause this is a disease that is caused through a mutation that happens in utero. These kids are born with this disease, and these lesions start to kinda grow with the child as they grow. They tend to find their way early on to pediatric hematologist oncologists in the more moderate, severe presentation. Great. Yeah, it was going to lead to my next question. It is sort of you recently started a pediatric cohort in patients, and that I guess makes sense from that standpoint because I guess the breast cancer majority of diagnosis probably happens when these children are still young. That's right. Yeah. maybe for the data that you're going to be disclosing in the first half, what type of data disclosures that we could expect from that, the cohorts of patients that you're evaluating? Yeah. Just to maybe level set the study we're running. Mm-hmm. Because of the breadth of experience we had in oncology, when we went into vascular anomalies, we were able to go into a dose randomization study as opposed to a traditional dose escalation. We're randomizing against three different doses of zovegalisib. The highest dose is our RP3D in oncology, so 400 milligrams BID, then 300 milligrams BID and 100 milligrams BID. So a good spectrum of target coverage. It'll be randomized, so about an equal number of patients across each of those doses. We said that the disclosure we'll make in the first half will have at least 20 patients that will be efficacy evaluable. What we mean there is they will have reached at least that first 12-week MRI time point. Nice. Safety database will be larger than that, but at least 20 patients of efficacy evaluable. It'll be largely PROS patients, the PIK3CA-related overgrowth spectrum patients, with a smattering of lymphatic malformations and venous malformation patients. Got it. When we think about I guess, potential pivotal design, assuming a success here in these early clinical trials, how should we think about that ultimately? The current precedent here from alpelisib, again, similar to the approach that we're taking where they took their non-selective molecule in oncology that they tried to apply here. That drug has accelerated approval right now. It was based off of a retrospective chart review of 37 patients treated under compassionate use. Kind of speaks to the level of unmet need in these patients. They failed their first confirmatory study, EPIK-P2, and they have a new confirmatory study ongoing. That is a single-arm study, 105 patients. We believe the accelerated approval pathway, depending on our data and interactions with the agency, would be available to us here. The bookends, if you will, of the 37 patients currently that gave alpelisib their accelerated approval and the confirmatory study of 105 patients. Mm-hmm. It'll be data and regulatory interaction dependent, but that's what's the precedent today. Yeah. I guess that first sort of missed study, is that they're attributing that of alpelisib to design or maybe the types of patients that they? Are there any learnings from that that you can use as you sort of think about your pivotal design? Yes. Their accelerated approval, there's no dose optimization 'cause it was just patients treated under compassionate use. They just took the oncology dose and backed off a little bit. They used 250 milligrams once daily. Because of that, when they went into the confirmatory study, the physicians on that study required them to go to half that dose 'cause they were concerned about safety and tolerability. They went from 250 down to 125 in the confirmatory study, saw about half the efficacy. Actually quite encouraging for us to see that 'cause we can see a clear dose relationship in terms of target coverage and dose intensity. Mm-hmm. We know from our oncology experience and everything we've done with the PK profiling against alpelisib, we can sustain much higher PI3K and target coverage than alpelisib can throughout the day. All of these pieces give us a lot of confidence as we move into our study that we should be able to demonstrate our superior efficacy with a better tolerability profile. Have you had conversations with physicians around their experience in the alpelisib, and then ultimately what that could translate for you? We've talked to physicians globally about their experience with alpelisib. We've seen there's been some publications, a number of case studies coming out of these academic institutions. It's clear that there's a tolerability profile issue on the table for what is limiting the full potential of the target. Great. maybe taking a step back some of the runway conversations that you've had with cash, and ultimately when you think about sort of the studies that you have going on in lymphatic malformation as well as breast cancer, ultimately, where did that cash take you from sort of a readout across these two pillars? Yeah. I mean, what we did over the last few years has taken our cash into 2029. That obviously fully contemplates executing the second line metastatic breast cancer trial. it goes a long way towards obviously, we don't know the exact trial we will run yet in the vascular anomalies, but it should take us a long way towards getting that towards the top line data. Mm-hmm. Obviously, we have money for moving our NRAS and Fabry's assets forward too, as well as running research. All of that is contemplated in the current cash that we have. The thing that's kinda sitting outside the cash window that we have is obviously running this frontline trial. We believe we have a lot of catalysts ahead of us the three that we've talked about and the data that we're about to show. We'll have ongoing updates from all three of these areas. We do believe that we have a lot of levers as we start to think about how do we path from here towards having three very large commercial opportunities in three different Phase threes all executed. Hopefully, we'll have revenue in the back half of this decade. We think we sit in a very nice position as we kind of execute all these opportunities. Great. Looks like we're up on our time, Sanjiv. Peter, thank you so much for your participation. Thank you for our listeners, and we'll be back soon with our next session. Thank you. Great. Thanks very much.