Okay, let's get started. Hey everyone, thanks so much for attending the Jefferies Healthcare Conference. Really appreciate you guys being here so bright and early. I'm Amy Li, I'm a biotech analyst, and I have the pleasure of welcoming the Relay Therapeutics management team. We have Sanjiv Patel, the Chief Executive Officer, Don Bergstrom, the President of R&D, and Pete Rahmer, the Chief Corporate Development Officer. Relay is a stock that we've been consistently flagging as incredibly interesting at these levels. They have a PI3Kα inhibitor in breast cancer. You're seeing at SABC, and especially at ASCO, that the space is heating up, and you're seeing signs that this mechanism can move up lines of therapy. They also have a very intriguing opportunity in vascular malformations, which currently has nothing approved. With that, I'd love to turn it over to the Relay team for opening remarks.
Thank you, Amy, and thank you to Jefferies for the invitation. I think we're excited here. We've just come back from Chicago, where we showed our own data for RLY-2608, in the second line and second line plus settings, where we were excited to see our data now really start to look very consistent and solidify, at an 11-month PFS for patients in the second line in combination with fulvestrant, and then a 39% response rate. Those numbers are numbers that we've not seen before in the class. As Amy said, the space, you know, if you looked at kind of a couple of years ago now, there was a lot of hope for PI3Kα mutated patients, and their treatment was hopefully at that point going to get better. This is not a small patient population.
It's 40% of all hormone receptor positive, HER2 negative breast cancer patients have a PI3Kα mutation. It's truncal and foundational, and it doesn't appear through resistance. It's there very early in disease. If you looked a couple of years ago, there were various mechanisms of action that were being tested, and, you know, even CDK4/6 retreatment. No one's been able to move the bar from kind of five and a half to seven months' worth of PFS, driven because the best way to treat a PI3Kα mutation is a selective PI3Kα inhibitor. I think over the last few years we've seen a couple of things. We've obviously seen validation of the mechanism in the second line with the approval of alpelisib. We've seen that if you can just get a better tolerability profile, people will use it.
That is what we see with the AKT inhibitor capivasertib, and it is doing remarkably well in its launch. Not because of its efficacy, it is all driven by the fact that it has a potentially better safety profile. Now, unfortunately, in the real world, we do not think that is playing out with the rash and diarrhea. I think the final thing to note is obviously the great work Roche has done over the last few years showing now overall survival benefit for a PI3Kα-driven triplet in the front line. It sets up a very exciting future for the mechanism. Obviously, we are the first selective PI3Kα inhibitor in the clinic, and we are now racing towards starting a pivotal trial in the post-CDK4/6 population. It is an exciting time for the field, and we are excited to be here today.
Awesome. Just since you came straight from ASCO, can you talk to us about what in your data that you found most encouraging, and also where do you think the field is headed for this mechanism? And any, and then any early feedback from physicians on your data?
Yes, absolutely. Maybe we will take those one at a time. The solidity of the data is what we are excited about. Eleven months PFS now in patients that have over one year's worth of follow-up, and a 39% response rate. In the kinase-only subset, we are seeing a 67% response rate and well over a year of PFS. I think we have just never seen numbers like this. Obviously these are 2x what we see for capivasertib for PI3Kα mutated patients of five and a half months. In terms of the field, I think we have talked a little bit about it, which is, you know, obviously we have seen a little bit of the kind of excitement of the other mechanisms start to come away.
Obviously the oral SERDs, I think the data showed from VariTAC2 and Serena 6, that they are going to be effective most likely only in the SR1 mutant population. Obviously those are resistant mutations that come in later lines. We saw the CAT6 data, as well. Again, patients with PI3Kα mutations do not do as well. Obviously there is a heavy toxicity burden that goes with those mechanisms. I think if you want to treat a PI3Kα mutation, use a selective PI3Kα inhibitor. The safety profile is the thing that really got the PIs excited, in Chicago, which is, you know, they deal with these patients day to day. Obviously a lot is said about grade three hyperglycemia, which, you know, doctors do not want to treat. These patients are hospitalized and require insulin.
I think what we're hearing more and more is the diarrhea is debilitating. That diarrhea is, you know, very clearly present with capivasertib, alpelisib, inavolisib. Even more troublesome is the rash. I think what we're excited to show is a very clean safety profile, very low levels of grade three hyperglycemia, diarrhea, rash, stomatitis. If you stack those two together, a nice safety profile with an efficacy profile that we've never seen, I think this bodes very well for the post-CDK4/6 population. I think we're also now thinking after the Roche data, which is a nice proof of concept. Roche, for those of you who follow the field, with their non-selective PI3Kα inhibitor, showed overall survival benefit in the front line endocrine resistant population.
You know, that to our mind, that, you know, they've proven out that it's possible. Now if you could just actually keep patients on this therapy with a safer profile, and like I think RLY-2608 in our mind looks like it has the potential to do that, you can service a very large patient population.
Excellent. Just talking about the Roche data, so they showed, like you said, an overall survival benefit. They showed a progression-free survival of 15 months. And, you know, there was a lot of talks and a lot of, you know, hyperglycemia, right? So can you talk to us about how much the therapeutic window of 2608 could improve the durability on survival by? And what would you consider, you know, as a win, for your front line triplet combo?
I think when we look at the INAVO120 data, that was the trial that you cited that showed a 15-month PFS for inavolisib in triplet, so fulvestrant, palbociclib, and inavolisib versus fulvestrant, palbociclib. You know, I think the toxicity profile is significant. As we talk to physicians and ask them which patients they're using this regimen in, they're either not using it, or they're using it in a very, very heavily selected patient population. I think Roche, wisely from a drug development perspective, as they were developing this regimen, went into a patient population specifically with regard to their metabolic baseline characteristics who were very, very fit patients. They largely excluded prediabetic patients. They required a baseline hemoglobin A1C less than 6%. They required an absolutely normal fasting plasma glucose at baseline.
What that translated into when you look at their trial is it was a very, very young trial. I think 80% of the patients were younger than 65 years of age. It was a very fit patient population. Despite that, and despite fairly aggressive use of prophylactic medications, including metformin for hyperglycemia, antihistamines for rash, mouthwashes for stomatitis, they still saw significant rates of those AEs. You know, I think as we are thinking about going into front line therapy, we are thinking about what is the best regimen to really be able to maximize therapeutic window for patients and to use a regimen that would be relevant in the world we are in today. I think one of the criticisms of INAVO120 is using the fulvestrant, palbociclib comparator at seven and a half months PFS is not a relevant comparator.
Right.
When a ribociclib doublet, just a doublet, can get you the same 15 months that you're getting with the inavolisib triplet. I think as we move into front line settings, we think that the profile we're establishing for 2608 gives us now the opportunity to significantly improve upon the risk-benefit profile you see in a triplet compared to what you see with inavolisib, and gives us the optionality to consider regimens that will be more relevant in the world moving forward than a palbociclib backbone.
Awesome. And, in your late line data at ASCO, you've effectively doubled the PFS of what we've seen with Piqray and some of these non-selective approaches. How do you expect that benefit, on durability, to translate in a front line setting? Like you said, that precombatives palbo, you guys are running it with a TIRMO. We've actually seen really interesting data that suggest the TIRMO might be better in terms of overall, you know, ORR, as well as some of these AEs, compared to palbo, right? Talk to us about, one, how did you select the regimen? And two, what, how do you kind of expect that PFS benefit to translate in earlier lines?
I mean, I'll start, maybe Don, you can take it, which is obviously the mechanisms here have all been validated now. Treating a PI3Kα inhibitor, the Roche data has shown us, can lead to greater benefit. I think if we could just continue to keep a very high coverage level on in the dose intensity, which we've shown in later line patients in the front line, I think you would see the efficacy that, that Inavolisib hasn't been able to achieve. A big part of that is you're stacking different drugs on top of each other, and it is the compounding toxicity that has been a challenge in these combinations. That is why obviously you see, you know, a significant burden of toxicity in the CDK4/6 inhibitors, which hopefully, CDK4 selective inhibitors will reduce.
If you stack a less toxic profile of a selective CDK4 on a much less toxic profile of a selective PI3K, hopefully you can keep these patients on therapy for longer. Obviously, if you're achieving the coverage that we can get with RLY-2608, it should translate into greater efficacy over time. I think the mechanisms have all been validated now. I think we just need to go ahead and run these trials.
Right.
Yeah. I just add to that that one of the challenges with the current generation of PI3K pathway inhibitors, whether it's alpelisib, inavolisib, capivasertib, is actually maintaining patients on drug, right? It stands to reason that if you can't maintain dose intensity or dosing, you're going to compromise efficacy. What we've continuously seen across this class, whether it be alpelisib, capivasertib, is a high rate of discontinuations due to AEs and generally a duration of exposure to drug in clinical trials that is several months shorter than the duration of the progression-free survival, indicating that patients are actually discontinuing exposure to the drug prior to the emergence of disease progression. What we've seen so far in later line patients, and again with now over a year of follow-up for 2608, is we are maintaining dose intensity.
We're at 92% median dose intensity, very, very low rate of discontinuations due to AEs. In a patient population overall, where you look across all lines of therapy, we have a median of 10.3 months PFS. We have a median duration of exposure of about 10.3 months, about 10 months. Patients are staying on our drug until disease progression, not coming off because of the safety profile of the drug. We anticipate as you move into earlier line regimens, this will become even more important, especially when you start talking about endocrine sensitive patients where you're talking about now treating patients for two years or longer.
Awesome. And Don, you made a, made a really interesting point earlier. inavolisib, the triplet combo, the trial enrolled a lot of healthy patients. You can probably assume that these are the types of patients that could tolerate this regimen in the real world setting as well. I’d love it if you can talk to us about how you designed your triplet in terms of the patient profile that you’re trying to enroll and why you think you could have a broader uptake than what Roche is showing right now.
We are currently still establishing doses with triplets in the context of our ongoing ReDiscover One trial. In our ongoing ReDiscover One trial, we are using metabolic inclusion criteria that are much more reflective of a general Western patient population. We allow patients to come on with baseline hemoglobin A1C up to 7%. That includes patients who are prediabetic. We allow patients to come on with a fasting plasma glucose up to 140. That includes patients who have some dysregulation in their fasting plasma glucose. We are using those inclusion criteria in our phase III trial, which we call ReDiscover Two. Again, generally representative of a, you know, Western patient population. We anticipate that as we would continue front line development, we would continue to use similar inclusion criteria.
Great. Moving on to, I mean, we've seen that Lilly recently acquired a competitor. I, and, I think there's been kind of questions around the, you know, equivalent exposure. So, yeah, tell us, in terms of the PK work that you've done, how, how much IC90 coverage are you getting with your 400 mg twice a day dose that is going to be the go-forward dose? And, do you think there is a coverage difference between you and the competitor drug? And how do we compare the data on an apples to apples basis?
You want to start with the PK?
Yeah. At the 400 milligram BID fed dose, which will be taken into the phase III study, we estimate we're sitting probably on average at around the IC90, maybe just below it. For all patients, we're seeing a clear coverage above the IC80. On average, just below the IC90, but for all patients, when you look at our PK curves, the bottom of the error bars are staying above the IC80. You know, as we synthesized STX478 and just ran that experiment the same way we run the experiment to determine target coverage with our own molecule, we see that it's not until you get to the 100 mg once a day dose of STX478 that it starts to come close to approximating the 400 mg BID fed target coverage. That's, you know, the fact is we see it.
We ran that experiment the same way you've seen us run it and published in Cancer Discovery. It's not until the highest dose tested, that of 478, that you see similar target coverage. I don't think we're certain what dose they'll actually be able to end up take forward at this point.
Got it. Got it. What would you say if investors asked you, said that maybe you were dosing lower than the competitor drug? Can you walk us through, I guess, how you're thinking about your PD markers as well as the data that you've seen in comparison to the competitor drug so far?
I would say to ask them to generate the data we've generated. I mean, that's how you compare these more accurately. We have generated clear, distinct data, 11-month PFS in the second line, 40% response rate, 67% response rate in county patients, up to 18 months of median PFS in those patients. With more than a year's worth of follow-up, you know, the best data we've seen from them publicly is a month and a half worth of follow-up on 50 or so patients in a monotherapy setting. There's not a lot of actual comparisons to be made there from a clinical data standpoint.
Right.
I think in the comparisons that can be made, I think by investors, I think we'd point you to the safety. Obviously they showed seven weeks of safety data. I think the unfair comparison was made with our nine and a half or 12 months worth of safety data in the combinations. Clearly there's a difference. The closest data we have is back in 2023. We showed 16 weeks of follow-up, and the data shows pretty, even though it's twice as much follow-up as the STX compound, shows it's comparable. What we didn't see, and have never seen, is the LiverTox that they saw after seven weeks. We didn't reach a DLT in our dose escalation, where they saw, I think, myalgia and paresthesia. In terms of, you know, there's a lot being said around monotherapy efficacy comparison.
I think STX showed something in the mid-20s for their breast cancer response rate. I think we've only dosed in monotherapy like for like three patients, of which one of them saw a response. So it's hard to see that, you know, if we really focused on it, we would see anything different. I think what we can say is robustly, we have dosed in combination, you know, a variety of breast cancer patients. The ones that have had prior fulvestrant, where you imagine the fulvestrant isn't doing a whole lot, we're showing a 40% response rate. I think that's probably as robust a comparison as we can make with their kind of mid-20s numbers. If you look at the other way around, our combination response rate now is 39%. We think fulvestrant, you know, adds somewhere between 3%-5% from prior studies.
If you stack that on their mid-20s, there's quite a gap that they have to assume that fulvestrant is going to give them, to get close to us. And given they're behind us and we've seen in precision oncology, if you come behind a drug, you have to do much better. The fulvestrant really has to help them.
Right. That makes a lot of sense. We also saw DESTINY-Breast 06 data, which was interesting, right? Because they had one, they did extensive ctDNA testing in, and they're seeing, I think, around 17%, 17, sorry, 14% patients actually getting tested or treated with a PI3K inhibitor front line setting. And they're showing a, I think, 13.2 month PFS in a second line, second line plus setting. Talk to us about, the data was in HER2 low population, but you do see a good amount of overlap between low HER2 and no HER2 expression. Talk to us about how you see 2608 fitting into the evolving landscape. And from a testing perspective, have you seen PI3K, you know, testing improve with Roche? And how do you kind of envision that evolving over time as well?
Yeah. The first thing I point out is the DESTINY-Breast 06 trial, while they're molecularly, the patient population is overlapping, clinically the patient population is quite distinct. The DESTINY-Breast trial in HER2 low patients is in patients where generally chemotherapy would be the therapeutic choice that would be used. You look at the inclusion criteria, those are either patients who had already seen two prior lines of endocrine therapy plus or minus targeted therapy or patients who had seen one line, and had progression on their front line therapy within six months of initiating treatment. These are actually patients who generally would not be eligible for the trial that we are about to start, the phase III trial that we are about to start. It's a patient population with much more aggressive disease.
It's a small patient population, compared to the patient population that generally you'd be treating with an endocrine therapy based regimen. With that in mind, you know, I think it's not surprising that you wouldn't see a strong dependence on PIK3CA. I think it's clear that, you know, the ADC is superior to chemotherapy, but as we talk to investigators, and I think as we hear from the podium and in the discussions after these presentations, in general for patients with targetable mutations, going through lines of endocrine therapy plus the targeted therapy prior to moving to chemotherapy is the general treatment paradigm, with the exception again of these patients with very aggressive disease who otherwise would need chemotherapy. I think clinically it's a distinct population from where we are.
I think as we develop 2608 and as we build the database for 2608, we would continue to see even stronger evidence for maintaining patients on an endocrine, chemotherapy-free regimen prior to going to the ADC. I think it actually, you know, as we have talked with investigators, you know, across trials, including in the INAVO120 trial, you know, one of the analyses they found most interesting is the ability for the inclusion of the PI3K inhibitor to significantly delay time to initiation of the first chemotherapy. I think, you know, as we are building our database for 2608, one of the goals in this patient population is to actually hold off using chemotherapy. ADC is just chemotherapy.
Right.
Right. Using chemotherapy as long as possible. I think with agents emerging like 2608, with the profile that 2608 will have, it both, you know, will satisfy that need, right, to be able to not only extend PFS, but to delay the use, or the time to first chemotherapy and further drive uptake of PI3K testing.
Excellent. I want to move on to vascular malformations, which is an equally as, potentially as big opportunity, that's currently underappreciated. I guess starting off, can you talk to us about the populations you're targeting, the opportunity that you see in this population? What is the current standard of care and where you see the big opportunity?
Yeah, sure. Just to level set, from a prevalence standpoint in the United States, there's about 170,000 PI3K, PIK3CA mutation-driven vascular malformations. Now, vascular malformations is an umbrella term of a handful of subsets of diseases. The one that has a current accelerated approval of a systemic therapy, which is alpelisib, is a subset called PIK3CA-related overgrowth spectrum. The entirety of that population has PIK3CA mutant-driven disease. There's about 5,000-15,000 of those in the United States. They do tend to have more severe disease because, as the name suggests, it is a spectrum of manifestations of the disease that present clinically. They tend to be more severe than some of the other subtypes.
The next, largest subtype, as a percentage frequency driven by PIK3CA mutations is lymphatic malformations. There are about 80,000 lymphatic malformations from a prevalence standpoint in the U.S. and about 80% of them are characterized as the driver of the disease being PIK3CA mutations. The next two would be venous malformations, 20%-25% have PIK3CA-driven disease, and then cerebral cavernous malformations, a little bit lower than that from a frequency standpoint. Our first-in-human study will have more of a focus towards PROS, so that PIK3CA-related overgrowth spectrum and lymphatic malformations.
Okay. Awesome. For PROS and lymphatic malformations, what is the standard of care right now? Are patients, is there anything approved in it or is it mostly surgery or symptomatic management?
Yeah. The one thing to appreciate in this disease setting is that it's a very nascent disease. We only have been characterizing these malformations or vascular anomalies as such for about the past 15 years. Serolimus, a very old multikinase inhibitor, immunosuppressant, has been used off label for about the past 10 years or so in this patient population. There is no formal label for serolimus in this, for these patients, but for severe patients, from what we understand, it is a systemic treatment physicians reach for. In 2022, Novartis got approval for their non-selective PI3K inhibitor, alpelisib. It's marketed as Vijoice in this setting. These patients exist on a spectrum of severity.
I think that's something we're all still learning, to understanding the distribution of the severity spectrum for these patients and exactly what they are seeking to treat their disease. There's some that are just watch and wait, and so they have a very single focused manifestation of the disease as a malformation on the skin in one geography to multifocal disease that could, the growth could be impeding other organs. I think one of the things we'll be learning over time is exactly the answer to this question, what is the spectrum of disease severity, and then what are patients that are best suited to contemplate chronic systemic therapy for that.
Awesome. Vijoice had, I believe, the EPIC study that showed, I think the data was interesting. It still showed hyperglycemia. Talk to us about, I guess, the treatment paradigm in PROS compared to kind of breast cancer, right? Is the benefit for 2608 going to be, you know, the durability as you see with breast cancer, or could it be you're hitting these mutations harder?
The concept is similar. You know, the goal of targeted therapy, both safety and efficacy are kind of two sides of the same coin.
Right.
you know, as in oncology, we are able to, widen the therapeutic window. Does come greater efficacy for these patients. The label dose for alpelisib in this setting is 250 mg once daily based off of the accelerated approval. And we saw a response rate in the high 20s. When they went to the confirmatory study, they dropped the dose in half to 125 mg once daily for adults. We saw the efficacy come down by almost half.
Interesting.
It gave us the confidence that this hypothesis is likely going to play out the way we anticipate it will, whereas we are able to push that dose intensity, give these patients better target coverage. It should ultimately result in better efficacy for them in the backdrop of a more tolerable profile.
Okay. Awesome. In terms of regulatory path forward, will you have an opportunity for accelerated approval in some of these populations? How are you envisioning designing this study?
There's currently no fully approved molecule in this setting.
Right.
You know, hesitant to say definitively that there's an accelerated approval pathway. That is a decision made by the regulatory authorities. As it stands today, that path is available to us and we will generate the data and have conversations with the regulatory authorities to see if that is a possibility for us. Obviously we need the data first.
Right.
The trial we are designing would enable us to do that swiftly. It is a part one, two, and three fully inclusive protocol where we can move from randomized dose selection into expansions and, if necessary, a randomized, placebo-controlled study. The current regulatory status would suggest that there's the ability based off of our data to have an accelerated approval pathway.
Awesome. I guess lastly, have you disclosed your dose for this study?
Yeah. The benefit of having all the data we have in oncology is it allows us instead of doing the traditional dose escalation, we can go into randomized dose selection. We will start at the phase III dose for oncology, so 400 mg b.i.d. twice daily. We will go to biologically active doses below that in a randomized fashion.
Awesome. I think we're out of time. Thank you so much. Really appreciate it.
Thank you.
Thanks.