Relay Therapeutics, Inc. (RLAY) Earnings Call Transcript

Good day, ladies and gentlemen, and welcome to the Relay Therapeutics corporate update call RLY-2608 data presentation at AACR Annual Meeting. As a reminder, this conference call is being recorded. I would now like to introduce your host for today's conference, Mr. Pete Rahmer, Chief Corporate Development Officer at Relay Therapeutics. Sir, you may begin.

Thank you, operator, and good afternoon, everyone. You can access the press release from today, the slides we are reviewing and a replay of this call by going to the Investor Relations section of our website. As a reminder, during this call, Relay Therapeutics will make certain statements that are considered forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995, including expressed or implied statements regarding our strategy, business plans, objectives, the expected therapeutic and clinical benefits of our product candidates, the potential of our platform and our candidates and progress and timing and execution of our clinical trials. Such forward-looking statements are not guarantees of future performance, and therefore, you should not put undue reliance upon them. These statements are subject to numerous risks and uncertainties that could cause actual results to differ materially from what we expect. I refer you to our SEC filings available at sec.gov or on our website for a discussion of the risk factors that could cause our actual results to differ materially from those discussed here today. The forward-looking statements in this presentation speak only as of the original date of this presentation, and we undertake no obligation to update them or revise any of these statements. Today on the call with me are Sanjiv Patel, our President and CEO; and Don Bergstrom, our President of R&D. And with that, I'll turn the call over to Sanjiv.

Thank you, Pete, and thank you for those joining the call today. As you know, we shared this morning first-in-human data for RLY-2608 at the AACR meeting. As we previously stated, our unequivocal goal for this program is to deliver a step change in efficacy for patients with PI3K-alpha mutations. The data presented today are clearly a promising first step in delivering against our goal. In data shown we've have been able to meet or exceed the target PK/PD threshold while dialing down the primary off-target toxicities associated with this class, namely hyperglycemia, rash and diarrhea. This starts to demonstrate that selective inhibition of PI3K-alpha mutant, which has long alluded our industry can be achieved with RLY-2608. These data also show early antitumor activity across a broad range of PI3K-alpha mutations and doses highlighted by a confirmed partial response to the breast cancer patients with 12 prior lines of treatment on RLY-2608 monotherapy. These data showed the clinical proof of mechanism, which is what we hope to achieve with this initial disclosure. With the favorable exposures we can reach and the emerging differentiated tolerability profile we've shown today, we will now move forward in the second half of this year and open expansion cohorts to drive towards interpretable efficacy data. In addition to this, we will continue to advance our next PI3K-alpha inhibitor RLY-5836 as well as our CDK2 inhibitor and our ER alpha Degrader into the clinic. We look forward to building a robust business with a broad range of assets over the coming years that can address the broad opportunity in front of us. We believe this is a very important day for patients with PI3K-alpha mutant tumors. Before Don takes you through the details of the data, let's summarize how we got here. When were launched in 2016, we were one of the first of the new breed of biotechs in perception of leading edge computational and experimental technologies with the goal of making medicines both more efficiently and more effectively. We were initially focused on creating 4 molecules against clinically and genetically validated drivers of disease in oncology. Over these last 7 years, through our constant execution focus, we've shown that we can repeatedly design new molecules in-house, progress them into clinical development and validate our platform and approach with clinical data. As depicted on Slide 4, our Dynamo platform consists of integration of complementary emerging techniques and technologies spread right across both the computational and experimental disciplines, all netted together by our experienced scientists. Our goal is to create therapies with exquisite selectivity for disease-driving proteins that we're studying. This selectivity allows for greater target inhibition without being limited by off-target toxicities. All of this was exemplified at last fall's ESMO meeting when we shared early clinical data for our FGFR2 inhibitor, RLY-4008. Importantly, we believe this relationship can be generalized across our portfolio. By deploying our platform and strategy, we now have a broad range of precision medicine programs that you can see on Slide 5. We've moved from focusing only on small molecule inhibitors to now having degraders and chaperones in our platform. We've also expanded from oncology to pursuing programs in genetic disease, and we previously announced we have 3 clinical assets. And today, I'm excited to announce that our fourth clinical program, RLY-5836, has also entered the clinic. All 4 of these clinical programs have been entirely created in-house by our team and platform. As highlighted on Slide 6, today's focus is our growing breast cancer business and it's all anchored around our franchise of PI3K-alpha mutant selective molecules and emerge from our unparalleled knowledge of this target. The first of these programs is RLY-2608. And following its creation, we continued our investment in research and in development for this target, and our clinical experience will help inform the best approaches for massively addressing this bold opportunity and how best we will deploy our broad franchise and program. As shown on Slide 7, globally, there continues to be a significant unmet need in breast cancer with 196,000 new hormone receptor-positive, HER2-negative breast cancer patients being diagnosed each year in the U.S. alone. Specifically, there remain limited treatment operations for metastatic breast cancer in patients who have already been exposed to CDK4/6 inhibitors with progression-free survival in the second line of best available therapies plateauing around 7 months. PI3K-alpha alterations are an important therapeutic target that could address this and then over time become effective in earlier lines of therapy. Moving to Slide 8. PI3K-alpha mutations are the most common kinase mutation across solid tumors. There are about 100,000 breast cancer patients with a PI3K-alpha mutation in the U.S. annually, which clearly represents a significant therapeutic opportunity. Oncogenic alterations occur right across the [ genes ] with mutations in the helical and kinase domains being the most common. The relative size of this opportunity is larger when for mutations in genes such as EGFR and ALK which have created multibillion-dollar commercial therapeutic franchises. Normal tissue PI3K-alpha regulates glucose homeostasis. Consequently, approved a late-stage nonselective inhibitors such as alpelisib and inavolisib inhibit normal tissue PI3K-alpha. And clearly, this disrupts glucose homeostasis. The lack at selectivity of these agents also leads to additional toxicities such as rash and diarrhea. All of these off-target toxicities mean that today, there remains a high unmet medical need for mutant selective inhibitors and a significant commercial opportunity for us to address. You can see on Slide 9, as a result of the current approved therapies' lack of PI3K-alpha selectivity, target inhibition is limited by high rates of treatment interruption, dose reduction and discontinuation due to toxicity. Subsequently, the efficacy of the approved and late-stage therapies is also quite limited. Our goal in developing RLY-2608, which is the first allosteric mutant isoform-selective inhibitor of PI3K-alpha was to try to address these limitations. Moving to Slide 10. To do this, we started by solving the first full length structure of PI3K-alpha in contact through its regulatory subunits for both wild type and mutated forms of the protein. We then use these data to perform long-time scale in silica molecular dynamic simulations to identify potential druggable differences in the proteins. And we found a novel [ turnable ] allosteric pocket in the mutant protein that we thought could provide us an opportunity to develop a pan-mutant selective allosteric inhibitor. We believe these insights are unparalleled in the field and culminated in the discovery of RLY-2608. Moving to Slide 11. We'll now take a look at RLY-2608's development in more detail. Slide 12 highlights the biochemical data validating 2608's unique mechanism of action. As you can see, it selectively targets mutant PI3K-alpha in contrast to the allosteric inhibitors that potently target both mutant and wild type. The kinome plot further shows 2608's exquisite selectivity right across the PI3K family as well as, in fact, the entire [ climate. ] On Slide 13, you can see further validation of 2608's mechanism in mouse breast cancer models. The graph show 2608's robust dose-dependent efficacy in both helical and kinase models. And the 2608 has little or no impact on glucose homeostasis with efficacious doses. In contrast, the doses required for efficacy, which are actually not clinically achievable, the orthosteric inhibitors significantly disrupt glucose homeostasis. As you can see on Slide 14, 2608 selectivity for mutant PI3K-alpha has been demonstrated in preclinical models. Meaningful target coverage and inhibition of the pathway does not lead to off-target toxicities, including most notably, glucose disregulation. All of this ultimately demonstrates tumor regressions in preclinical models. Moving to Slide 15. We highlight our initial objective for today's disclosure, which is to demonstrate clinical proof of mechanism. This is similar to the focus of our first disclosure for our FGFR2 inhibitor, RLY-4008. What this translates to is showing early and initial PK and PD data that supports the fact that we can engage the target at or above levels associated with efficacy in preclinical models. It also translates into showing early safety and tolerability data to show that we can avoid key off-target toxicities at these high levels of target inhibition. And this is exactly the data that we'll show you today. Moving forward, the goal of future disclosures is to generate data that will allow us to further evaluate the potential efficacy profile of RLY-2608. Moving to Slide 16. Our unequivocal goal for this program is to deliver a step-change in efficacy for patients with PI3K-alpha mutations. The data presented at the AACR meeting today, our first step in delivering against this goal by demonstrating selective targeting of mutant PI3K-alpha These data show we can achieve selective mutant PI3K-alpha target inhibition above the IC80 threshold associated with efficacy in preclinical models. We can do this with limited observed impact on glucose homeostasis and no grade-3 hyperglycemia being observed. All of this clearly speaks to the fact that 2608 is selective over wild type. This selectivity translates into an overall favorable safety profile. Adverse effects are low grade and generally well tolerated with no dose-limiting toxicities or adverse events leading to treatment discontinuation. The therapeutically active doses, we observed low rates of rash, diarrhea and most notably hyperglycemia. All of this translates into a potential greater dose intensity for 2608. And this is manifested in early antitumor activity that we observed across a range of doses and across a range of PI3K-alpha mutations. All of this, as you know, has also been confirmed with our partial response being observed in a heavily pretreated breast cancer patient that was on RLY-2608 monotherapy. The totality of these early clinical data has started to demonstrate our goal, which is to show clinical proof of mechanism for this program. We look forward to opening the expansion cohorts and generating robust and inseverable efficacy data. All of this, we believe, brings us a step closer to potentially realizing our goal of developing a medicine that could provide a step-change in efficacy for patients. And now I'll hand it over to Don Bergstrom to address the data in more detail.

Thank you, Sanjiv. Starting on Slide 17, ReDiscover into dose escalation, dose expansion study with key objectives to define the maximum tolerated dose and recommended Phase II dose [indiscernible] CDK and antitumor activity of RLY-2608 alone and in combination with fulvestrant. The monotherapy arm focuses on PI3K-alpha mutant solid tumor patients, including breast cancer patients, not eligible for the combination arm. The combination arm focuses on hormone receptor positive HER2-negative breast cancer patients treated with prior CDK4/6 and endocrine therapies. Today, we'll present initial results of the dose escalation portions, both of which still are ongoing and open to enrollment. The cutoff date for the data in the presentation was March 9, 2023. As shown on Slide 18, 12 dose levels on a continuous daily dosing schedule were examined in these dose escalation using a bayesian optimal interval design with additional accrual permitted to cohorts declared tolerable. Across both arms, maximum tolerated dose was not reached with no DLTs observed. Dose escalation in cohort enrichment continue to define the recommended dose or doses to take into cohort expansions. Slide 19 summarizes the study population. 42 patients with PI3K-alpha mutations were enrolled with 19 treated on the monotherapy arm and 23 treated on the breast cancer combination arm. Both populations had significant prior therapy. In the monotherapy arm, the median number of prior regimens in the metastatic setting was 4, with the majority of patients treated with prior chemotherapy or an antibody drug conjugate. This arm included a broad diversity of solid tumors, predominantly tumor types not predicted to be sensitive to single-agent PI3K-alpha inhibition. In the combination arm, all patients were treated with prior endocrine therapy and the CDK4/6 inhibitor per protocol. About half the patients in the combination arm had 2 or more prior systemic therapies for metastatic disease and about 1/4 of patients received prior chemotherapy or an ADC in the metastatic setting, including patients treated with trastuzumab deruxtecan. While the protocol prohibited prior treatment with the PI3K-alpha inhibitor, 17% of patients were previously treated with an mTOR or AKT inhibitor. Also note that while the protocol required good glucose control at baseline, enrollment was open to patients at risk of prediabetes and more than 1/4 of enrolled patients had a BMI greater than 30. On Slide 20, we show the molecular demography of the 42 enrolled patients, a wide spectrum of oncogenic PIK3CA mutations were detected, reflecting the published prevalence of mutations of PI3K in solid tumors, most commonly seen were mutations in the helical and kinase domains. And 5 of the 42 patients had 2 or more detectable PIK3CA mutations. Slide 21 shows RLY-2608 pharmacokinetics for both study arms, with dose levels depicted on the left of the Y axis. The upper dashed line in both clots indicates the exposure associated with 80% inhibition of pAKT and associated monotherapy efficacy of 2608 in preclinical models. Note, favorable PK was observed across both arms with dose-dependent increase in exposure and low peak-to-trough fluctuation across dose levels. At trough concentration, the 400-milligram monotherapy dose achieved the predicted efficacious exposure associated with 80% inhibition of phospho-AKT, as shown in the left panel. The 600- and 800-milligram combination doses achieved continuous 80% inhibition of phospho-AKT for the duration of the dosing interval as shown on the right panel. Moving to Slide 22. To assess pathway suppression with patient samples, we developed an ex vivo pharmacodynamic assay using plasma samples collected at trough from each patient to assess suppression of phospho-AKT in PIK3CA mutant cancer cells. This graph shows the relationship between RLY-2608 at a trough PK draw and the extensive phospho-AKT inhibition measured at that time point. As shown in the graph, across dose levels, nearly all patients achieved at least 50% pAKT suppression in the trough plasma sample. pAKT suppression was concentration-dependent with robust 80% suppression at concentrations achieved by 400-milligram BID monotherapy and 600 and 800-milligram BID in combination with fulvestrant. On Slide 23, you can see that glucose homeostasis was generally preserved across all dose levels evaluated in the ReDiscover trial with a low rate of grade 1 or 2 hyperglycemia and no observed grade 3 hyperglycemia across all 42 patients. On Slide 24, we show glucose levels versus time for the dose cohorts that exceed 80% inhibition of phospho-AKT continuously over the dosing interval. The data represents the mean per cohort plus or minus standard deviation. Across these predicted efficacious dose levels, glucose levels remain largely within the normal range. Modest glucose elevation was observed in some patients, but this was readily managed with oral antihyperglycemic agents. No patient required insulin or experienced grade 3 or higher glucose elevation. These data, together with the PK/PD data show that 2608 has minimal to no impact on glucose homeostasis and exposures in the predictive therapeutic dose range. This supports mutant selective targeting of PI3K-alpha by 2608. On Slide 25, we show the therapeutic potential of mutant selective PI3K-alpha inhibition. The vignette shows radiographic response in a heavily pretreated woman who received 2608 monotherapy at the 400-milligram BID dose. She had received 12 prior lines of therapy for her metastatic disease, including Enhertu. As shown in the baseline scan on the left, the patient had high disease burden with multiple liver metastases. After 4 weeks of therapy, marked reductions of greater than 90% were observed for both detectable PIK3CA mutations as well as ESR1 mutant ctDNA as depicted in the middle panel. Subsequently, the patient achieved a partial response per RECIST with 36% tumor reduction at the first radiographic assessment as shown on the right. This response was subsequently confirmed after the data cutoff. The patient experienced no hyperglycemia or other adverse events due to study treatment, demonstrating mutant selective inhibition of PI3K-alpha by 2608. She currently remains on treatment. Of note, the overall monotherapy experience included only 4 patients from the indications we will focus on for our expansion cohorts, 2 head and neck patients and 2 breast cancer patients with more than 1 PI3K-alpha mutation. Of these 4 patients, only one was treated at a dose that met our target exposure threshold, and that was the patient who achieved a confirmed PR. On Slide 26, you see radiographic tumor regression and RECIST assessment for all breast cancer patients with measurable disease. The patients on the far right is the partial response highlighted in the previous slide. Overall, 56% of measurable patients had radiographic tumor reduction with stable disease or partial response per RECIST criteria with 11 of 16 patients ongoing on 2608. Progressions were observed across doses in PI3K-alpha genotypes, including patients with helical domain mutations, kinase domain mutations and other PI3K-alpha mutations. Moving to Slide 27. Declines in mutant PIK3CA ctDNAs are an indicator of pharmacodynamic inhibition of mutant PI3K-alpha. The graph show PIK3CA ctDNA levels at baseline and after 28 days of treatment with RLY-2608. Declines in PIK3CA mutant ctDNA were observed in most patients and occurred with both monotherapy and combination therapy and is emphasized by the abstract discussions, Dr. Fabrice Andre, when referring to the pharmacodynamic mechanism of RLY-2608. This is evidence that in his words, the drug works. The swim lane plot on Slide 28 shows duration of exposure in response for all RECIST evaluable breast cancer patients. 19 out of 27 patients shown here remain on study. Overall, the median duration of follow-up at the time of data cut was 16 weeks with a range of 4 to 44 weeks, including a number of patients treated at doses below the target exposure threshold who have remained on treatment for more than 6 months. Moving to Slide 29. This is the overall safety profile for RLY-2608 showing all treatment-emergent adverse events occurring at greater than a 15% frequency across monotherapy and combination therapy cohorts and all doses. There are no dose-limiting toxicities, AEs leading to discontinuation or grade 4 or 5 AEs. The majority of AEs were grade 1 or 2, manageable and reversible. Of note, the overall rate of hyperglycemia was low, and there was no grade 3 or worse hyperglycemia. Slide 30 shows a comparable distribution of AEs for RLY-2608 monotherapy on the left and combination therapy in the middle panel. The right panel shows AEs for the 17 patients who received doses that exceeded the target exposure for 2608 associated with sustained 80% inhibition of mutant PI3K-alpha. Of note, the overall rate of AEs in the target exposure population is largely reflective of the overall safety profile, including low rates of hyperglycemia, diarrhea and rash and no grade 3 events for these 3 key toxicities of nonselective inhibitors. Moving to Slide 31. In the SOLAR-1 trial of alpelisib, the 3 AEs that most commonly led to treatment discontinuation were hyperglycemia, diarrhea and rash. Toxicities such as these were treatment limiting as 25% of patients discontinued therapy due to AEs. High rates of hyperglycemia, diarrhea and rash, including grade 3 or worse events, have been observed for the nonmutant selective PI3K-alpha inhibitors, alpelisib and inavolisib. Capivasertib, a pan-AKT inhibitor demonstrate lower rates of hyperglycemia, but still demonstrates high rates of diarrhea and rash, including grade 3 or worse events, despite being administered on an intermittent 4 days on, 3 days off schedule. In contrast, as you can see on Slide 32, 2608 shows a low rate of the 3 AEs, most commonly leading to treatment discontinuation for alpelisib, including in the 17 patients who exceed the target exposure of RLY-2608. This toxicity profile in these data translates to a low rate of dose interruptions or reductions due to AEs and no treatment discontinuations due to AEs. And dose intensity is preserved with median dose intensity greater than 98% of the planned dose. Moving to Slide 33. In summary, as emphasized earlier today at the AACR meeting from both Dr. Varkaris and Dr. Andre, RLY-2608 has favorable PK/PD that provides robust target inhibition, including sustained inhibition of mutant PI3K-alpha at 80% or greater with limited observed impact on glucose homeostasis. We believe this initial clinical data has shown a favorable and differentiated safety profile consistent with mutant and isoform selective PI3K inhibition, which could enable treatment of broader patient populations and combinations that cannot be achieved with nonmutant selective inhibitors. RLY-2608 shows encouraging anti-tumor activity across PIK3CA genotypes and hormone receptor positive HER2-negative breast cancer. Together, we believe these proof of mechanism results demonstrate that RLY-2608 is the first allosteric and mutant and isoform selective PI3K-alpha inhibitor to enter the clinic. Moving to Slide 34. ReDiscover dose escalation continues to define go-forward doses for the planned monotherapy and combination therapy dose expansions, which we expect to begin in the second half of 2023. Moving to the next program in our breast cancer franchise on Slide 35. Our commitment to ensure that we deliver the best therapeutic solution for patients with PI3K-alpha mutated cancers is manifested by our decision in 2021 to bring RLY-5836 into development prior to seeing clinical data with 2608. I am pleased to say that the first-in-human trial of RLY-5836 has started. And as you can see on Slide 36, the design of the 5836 trial is similar to the 2608 ReDiscover trial. With that, I will now hand it over to Pete to wrap up.

Thanks, Don. RLY-2608 exists as part of a broader portfolio of breast cancer assets that we are advancing with the aspiration to bring about a step-change in treatment of breast cancer. The data we shared today validate the potential advantages of highly selective inhibition of validated breast cancer targets that lies at the heart of our breast cancer portfolio. As we transition to enrolling the expansion cohorts of patients in 2608 later this year, the goal is to demonstrate that this clinical profile translates into interpretable efficacy and a differentiated long-term safety profile relative to non-selective PI3K-alpha inhibitors. We continue to build on the foundation laid by 2608 with 5836, our second pan-mutant selective PI3K-alpha inhibitor now in clinical development. Also, our selective CDK2 inhibitor, RLY-2139, which is on track for entry into the clinic in 2024. And lastly, our oral bifunctional ER-alpha degrader on track for choosing a development candidate nomination sometime this year. In addition, our efforts against breast cancer targets continue with other undisclosed research stage programs. As you can see on Slide 38, we have a robust precision medicine pipeline with more than 13 active programs that continues to progress as planned and has the potential to address hundreds of thousands of patients annually in the U.S. alone. And we believe the data presented today continue to validate our approach. The more we do, the better we get. And lastly, moving to Slide 39. A key theme throughout the remainder of 2023 will be continued clinical execution particularly with our lead clinical programs. For 2608, we will continue initial dose escalation with a goal of starting dose expansion in the second half of the year. And as we mentioned earlier, we just initiated the 5836 first-in-human study. And we will make additional clinical data disclosures in 2024 and from either one or both of the pan-mutant selective PI3K-alpha inhibitor programs. We have 2 key anticipated milestones for RLY-4008 in the second half of 2023. First is the full enrollment of our pivotal cohort in FGFR2 fusion-positive cholangiocarcinoma. And then we will show initial clinical data from the ongoing non-cholangiocarcinoma expansion cohorts. With today's data, we continue to make progress toward our goal of delivering life-changing medicines to patients. We have an execution-focused team, sufficient capital and emerging track record to deliver on all of these milestones, and we look forward to updating you on our progress over the year. Before we turn the call over to the operator for questions, I'd like to ask all the folks asking questions, please limit to just one question per person. Operator?

[Operator Instructions] Our first question comes from Salveen Richter with Goldman Sachs.

Just maybe here on what you can do to optimize efficacy as the target engagement curve suggests that higher doses may not drive benefit. And you're sitting at kind of 600 and 800-milligram BID where we see PI3K inhibition plateauing at about 80%.

Yes. And maybe, Don, do you want to just take that one?

Yes. So I think that the number one thing we need to optimize efficacy is more time and experience at a given dose. Efficacy in this patient population takes time to establish whether it's measured by response rate, where for alpelisib in the Phase Ib trial with fulvestrant, the median time to response was 4 months or measured by clinical benefit rate, which is a 6-month endpoint. So I think we anticipate that we are seeing the degree of target inhibition that's been associated preclinically with seeing robust antitumor efficacy. We're seeing strong PD is manifested by the clearance of ctDNA, especially at the doses that we project to be in our efficacious range. In early scans, we're seeing evidence of tumor shrinkage. And we anticipate we just need to have more time and experience to be able to see those responses potentially continue. As well as to be able to estimate a clinical benefit rate, which ultimately, I think, will be our signpost as we're looking towards PFS, which is the ultimate registration endpoint in this patient population.

I think 6 out of 7 of our 600 milligram patients still remain on therapy. And so I think for us, it's the question is now of moving forward with a greater number of patients and a greater duration of follow-up. But I think all the things that we wanted to achieve, which is engaging the target, doing it with a cleaner safety profile we've been able to achieve. So I think we look forward to opening the expansion cohort to getting the data.

Our next question comes from Brad Canino with Stifel.

Congrats on the human-intended product profile here. On the dose management, I'm really struck for the lack of dose reductions, but most of the 800 mg BID patients are 8 weeks or less of treatment here at this data cut. So my question is, how does that compare to the appellate of experience? Do you see the majority of the of appellate of reductions within the first 2 months? And are we going to need to wait to see a longer follow-up to have confidence that you're avoiding that with 2608.

So maybe Don, just talk about the emergence of when the adverse effect emerge.

Yes. So I think the experience with alpelisib and I'll back up and say, I think the primary AE again, that led to reduction of alpelisib and the intensification of those things was hyperglycemia. And the onset of hyperglycemia in alpelisib is quite rapid, certainly within the first cycle frequently within the first 2 weeks of dosing. So I think the fact that now we have a number of patients treated at the 800-milligram dose that have been treated past that 2- to 4-week threshold, and we're getting more experience without seeing the onset of the hyperglycemia is reassuring that as we continue to get experience with a 600 and 800-milligram dose, we'll be able to maintain the dose intensity that we've seen to date. I'll also point out that by far, the vast majority of the AEs that we've seen across the profile for 2608 have been grade 1 AEs. They've not been clinically significant. And they have not impacted the ability to manage dose intensity and manage dosing of the patients.

Our next question comes from Jason Gerberry with Bank of America.

In the past, you guys have talked about a goal of getting to IC50 or lower coverage with the wild-type coverage. And so I didn't see that update. Maybe it's just inferred from the low rates of adverse events, but just curious if you can comment on the exposure to the wild type. And ultimately, how should we think about prioritization of 5836 after this data. Is it purely a hedge in the event of some unforeseeable idiosyncratic tox? Or is there anything more you'd add to that?

Great question. Maybe I'll take the first of the question, singular. 5836, so I think, as you know, when we shared that we had the program last summer, it was all around providing the best available treatment to patients. And we didn't know what would happen when you put a medicine -- you typically do see idiosyncratic things. And so 5836 has a different chemical structure, a slight different PK profile. And so it will -- as we announced on the call today, is in the clinic now. And we'll have both in the clinic for a period of time, where we will make this a choice on which to take forward. The IRA, as you know, provides us with an interesting opportunity to actually take both forward if we have to -- the indications. And so I think there is -- today, our need for 5836, seeing the proof of mechanism that we have on 2608 has probably gone down, the need for it. But there is just this one thing around the IRA that actually require us to take both forward in a certain situation. And to me with that, I'll bring it back to the first question around the IC50.

Right. So we didn't explicitly show IC50 data, but our goal using IC50 was using that as a threshold above which we were concerned we could run into wild-type PI3K-alpha toxicity. And I think with the tolerability profile that we've shown with low rates of Hyperglycemia, diarrhea and rash. And then again, generally, across all AEs, largely grade 1 AEs is consistent with us being in the range we wanted to be for minimizing the either wild-type PI3K-alpha related toxicity or toxicity related to other isoforms or family members.

Our next question comes from Yaron Werber with Cowen.

I got sort of interrelated 2 questions. I guess the first one, how much higher do you foresee going in terms of the dose escalation? And then secondly, when you do look at the -- I think it's Slide 26 at the measurable disease slide, the 16 patients. It looks like one arm actually crosses the 30 and that's not to mono, the combo, but it wasn't considered in the PR. Did they just not get scanned? Or can you give us -- expand on that.

Maybe, Don, you just take that.

Yes. So that was a patient who, again, consistent with the notion that responses come late for this mechanism in this patient population. This is a patient who had the cycle 6 scan, approximately 24 weeks into therapy, had shrinkage with target lesions down below 30% response. Unfortunately, at that scan, the patient did have detection of a new nontarget lesion. So a new lesion, which defines that scan not as a response but as a PD. Given the reduction in the tumor lesions that the patient had, the investigators felt strongly that this patient was responding from therapy, and this patient does continue on treatment with RLY-2608.

And then maybe, Pete, answer the question around how much dose escalate.

Yes. I think you highlight a good point, Yaron, which is -- we have a clear and discernible range of doses that hit the target profile of what we're looking to achieve. So we will continue to dose escalate to just to understand the full breadth of potential exposures and knowing that we're nowhere near having hit a DLT or MTD. But generally speaking, anything from the 400-milligram BID to 800-milligram BID looks like clearly a potential dose or doses to contemplate taking forward into expansion.

Maybe Don, just comment on the kind of dose intensity...

Yes. So I think, Yaron, as we go through looking at doses and selecting doses to take forward, one of the key things we'll be looking at is the ability to maintain dose intensity chronically. We've been able to show at the doses we have evaluated so far, we maintained 98% dose intensity. The reason why that's significant as we mentioned earlier, with alpelisib, you see significant challenges with maintaining dose intensity. And the average dose that's actually administered for alpelisib, the median dose that actually administered is about 250 milligrams daily compared to 300-milligram daily dose that is the label dose. And there is actually an analysis of the SOLAR-1 data that looked at PFS as a function of whether patients are able to stay above that median dose of 250-or-so milligrams or whether they're below that median dose. And there was a numerically significant difference in PFS for patients who are able to maintain dose intensity with alpelisib. So our goal as we choose a dose for 2608 is to choose a dose where we can not only hit PI3K-alpha hard as we are already, but maintain the type of favorable tolerability that allows you to maintain dose intensity. And I think based on that precedent experience with alpelisib, if you can maintain dose intensity that should translate into superior progression-free survival.

And so is there -- are you planning on dose-escalating further up? Or you're just planning on expanding the current doses?

So we can continue to dose escalate. And we do anticipate looking at higher doses, and we'll look at the totality of the data as we choose which dose or doses to take forward into expansion cohorts in the second half of this year.

Our next question comes from Michael Schmidt with Guggenheim Partners.

Is there something going on with the potential drug-drug interaction with fulvestrant? I'm just asking because you're reaching the IC80 in the monotherapy at 400, but it seems like you have to go a bit higher to 600 or 800 for the combination. And is that -- is there a signal perhaps then what would that mean for achieving higher exposures if we go higher in dose?