Good afternoon, and welcome to the Morgan Stanley Global Healthcare Conference. I'm Jeff Hung, one of the biotech analysts. For important disclosures, please see the Morgan Stanley Research Disclosure website at www.morganstanley.com/researchdisclosures. If you have any questions, please reach out to your Morgan Stanley sales representative. For this session, we have Repare Therapeutics with CEO Lloyd Segal and CSO Mike Zinda. Welcome, Lloyd and Mike. Thanks for having us here. Very excited to be here. For those who may not be so familiar with Repare, can you just provide a brief introduction? Sure. Repare today, as we imagined it to be six years ago, is the leading synthetic lethality-focused, DDR-focused biotechnology company out there. We have been, I think, tremendously successful in bringing our synthetic lethal compounds driven off our proprietary platform into the clinic, most recently with our camonsertib program partnered with Roche, which we'll present and talk about today. Following it on with proprietary and innovative new targets and compounds, including RP-6306, which is now a year into the clinic. We've recently identified two follow-ons to that we expect to be clinical over the next 18-24 months, starting with our Polθ compound. We have been incredibly prolific and have executed with pretty high quality against an ambitious goal. Maybe before diving into your pipeline, let's talk about your approach. You know, what is the SNIPRx platform, and what advantages does it provide Repare? Yeah, great. The SNIPRx platform is our CRISPR-based platform for identifying novel targets to treat patients with cancer, and in particular, to look for synthetic lethal interactions. To do that, we take a pair of cell lines, one that's wild type and one that has the genetic aberration of interest, and use that to interrogate the entire genome using a proprietary CRISPR library and identify targets that selectively kill that cell with a genetic aberration versus the wild type cell. Progress that through further validation, ultimately bringing those different targets that we identify into our drug discovery platform and then into the clinic to help treat patients. What are the step two screens and how do those expand the patient population? Yeah. For your programs? Yeah. The step two screens, again, are a screen using our CRISPR platform, but it was really building off of the idea that synthetic lethal interactions aren't necessarily just a gene-to-gene interaction, but can be a network of interactions, much like what was seen with PARP inhibitors that were beyond BRCA, but actually were being able to be impacted by HRD defects within the genome. It's a methodology to actually front-load that and prospectively identify those patients through using that CRISPR screen, and then be able to expand the patient population out that we are prospectively asking questions on clinically. Maybe moving on to RP-3500. In June, as you mentioned, you announced an agreement with Roche for camonsertib. Can you just talk about this deal and why was this the right time to partner this program? Yeah. I think as we outlined, a couple of months back when we announced the partnership, we realized as the data internally was coming to the fore, that the opportunity in ATR inhibition was even bigger and broader than we thought it was. We increasingly in 2020 came to a conclusion that finding the right partner who was committed as we were to winning in the battle for ATR inhibition. As you know, Bayer, AZ, Merck Serono, all had compounds that when we started were as much as three, four years ahead of us. That if we were really gonna have a chance to win and to prosecute the opportunities that the data were telling us were clearly there, that a partner was the right thing. We started a pretty systematic process led by our CBO, Kim Seth, and in concert with Mike and Maria and myself and Steve. It was a wonderful process. We learned a lot, but we came out the other end with a very clear picture of Roche's commitment to winning in ATR inhibition. That was the big driver. It was more important to us than any upfront or any economics. We really believe today, even more so than we did when we announced the deal, that Roche is completely committed to aggressive global development for the benefit of patients and all of our shareholders. Speaking of economics, you do have an opt-in to a 50/50 U.S. co-development and profit share. When is the deadline for the opt-in, and what factors will likely influence that decision? As you know, Jeff, and it's just typical with these kind of deals, a lot of that was redacted out for what Roche viewed as proprietary reasons. What we can say is it's a future event, and it's a trigger generated by some material clinical activity. It's in the truest sense an option. We'll make that decision at some point in the future, whether or not it makes sense for us to opt into what would be a 50/50 co-promote in the U.S., which then we'll have a lot of data that we don't have today. I think we'll have a better sense of how much that makes sense economically at the time as well. Okay. Well, let's move on to RP- 6306, PKMYT1 inhibitor. For those who may not be familiar with PKMYT1, can you talk about its role and what is the goal by targeting that? PKMYT1 was a synthetic lethal target that we identified using our proprietary SNIPRx platform, and found that it was synthetic lethal with cyclin E1, a key aberration in cancer, that was one of the earliest lesions that we linked through. I think it's a kinase that's involved in regulating the transition of cells into mitosis from S phase. Therefore, really we found was this great link between cyclin E, which was pushing cells into S phase early, driving genome instability, and PKMYT1, its partner, that was allowing those cells to actually be viable and progress forward. When you inhibited it, we saw this synthetic lethal interaction that really allowed us to think about a new therapeutic approach to cyclin E patients. What is CCNE1, and how does overexpression result in genome instability? Yeah. Cyclin E1 is a partner with CDK2, and it's responsible for progressing cells normally from the G1 phase to the S phase of the cell cycle. That would normally occur after the cell has already created all the intermediates that are necessary to complete DNA synthesis. But when cyclin E is amplified, cells exit G1 early, entering S phase before they have those building blocks. So now they're trying to create another strand of DNA, but they don't have all the building blocks necessary. That creates a lot of stress on the genome. It extends out the S phase and results in a very interesting interaction with PKMYT1 that we were able to identify. What has been the challenge for targeting CCNE1? Yeah. CCNE1 has been known for a long time, and I think the only target prior to our discovery of PKMYT1 that I think was of primary interest was CDK2, of which Blueprint and Pfizer have both found compounds, but the biggest issue there is I think, as most people know, CDK2 inhibitors have been looked for for about 25, 30 years, and really selectivity was the hardest thing to find there. I think there's some new inhibitors that are very interesting in that space now. This was a brand-new link that allows us to look at a interaction not directly with cyclin E, but the results of cyclin E's aberrations in those cells. How did you determine that PKMYT1 is synthetic lethal with CCNE1? Yeah. Again, as part of our SNIPRx platform, we created a number of different paired cell lines where we overexpressed cyclin E1 in those cells, showed that it created the same kind of alterations, i.e., pushing cells early into S phase, that they caused replication stress, so that there was genome instability. We took those pair of cell lines, we did a whole genome-wide screen, and then we compared and looked for hits that were selectively killing the cyclin E-amplified cells versus those that were wild type. PKMYT1 was one of the top targets that came out in that finding. We were also able to find that through the DepMap consortium, the work done by the Broad, and then through multiple levels of validation as well as a small molecule identification, where we then confirmed that with pharmacology, to make sure again that it wasn't a scaffolding or other component to PKMYT1 itself. PKMYT1 is a member of the WEE1 protein kinase family. How do you think about targeting PKMYT1 compared to targeting WEE1? You know, might there be any advantage on tolerability? I think the key thing that I would like to do is just step back and take a look at what are PKMYT1 and WEE1 doing? How are they similar, and how are they different? I think they both have one similar function, and that is to help control the cell's entry into mitosis. It's a very important phase of the cell. You need to make sure the DNA was replicated. The cell has created multiple methodologies to be able to put the brakes on the cells to not enter mitosis. Beyond that, they're actually very different. PKMYT1 is a cytoplasmic kinase. It's the first phosphorylation site that inhibits CDK1 in the cytoplasm that has to then be removed, and then CDK1 has to be activated, translocated to the nucleus, and then WEE1 can come in and shut it off again. Again, two very different sets of signals moving in there. Then the last part is PKMYT1, to the best of our knowledge in all the literature, is that it's really only regulating CDK1, whereas WEE1 also regulates, CDK2 and the G1/S transition. We think that's what was causing the differences that we saw when we looked at our synthetic lethal screens and when we followed them up with pharmacology. We saw around a 40-fold window of selectivity between the cells that had cyclin E overexpression versus the wild type, whereas for WEE1, we saw a very narrow window that was more around one to three fold in those settings. We think those two things are different. To your other point of, is there differences in how inhibiting PKMYT1 might manifest in toxicology? We haven't revealed the toxicology for a number of reasons, but what we have said is that it looks very different than inhibition of WEE1, ATR, CHEK1, or any of the cell cycle checkpoint inhibitors. We think it looks really favorable for being able to combine with other therapeutics to hopefully achieve maximal benefit to patients when we progress into the clinic. Why do you get better selectivity on CCNE1 overexpressing cells with RP-6306? Yeah. In a wild-type cell, cyclin B and CDK1 are coming up at a certain pace. By that time, by the time they're up and they're ready to be activated, the DNA has already been replicated, and PKMYT1 really doesn't matter in those cells. Whereas when cyclin E1 is amplified, because of that extension of the S phase, CDK1 and cyclin B still come up at the same rate, but now they're sitting there for a long period of time, and PKMYT1 becomes extremely essential to holding them at bay until that DNA has been replicated and it's time to go through mitosis. That's what gives that difference between a wild type cell and one that has cyclin E amplified. A lot of this, you know, that Mike's covered in the last few minutes, it was elucidated in Nature paper, I think you're familiar, which is available off our website, and really nicely captures this complex story, but understanding that differentiation is so important. Mm-hmm. What have you seen for combinations with 6306? Yeah. It seems some really exciting activity, and it's triggered three different combinations that we're doing in the clinic. One is with gemcitabine, which actually kind of furthers that cyclin E1 amplification phenotype. It helps deplete nucleotides that are necessary for DNA synthesis and puts extra stress and extra reliance on PKMYT1 for the cells. Irinotecan works in a slightly different way, where it's actually driving replication stress directly. Again, we're making the PKMYT1 reliance in those cells even further. The third one is a really exciting combination with RP-3500 camonsertib. That's more of a mechanistic base endpoint because ATR activates CHK1, which then activate or inactivates CDC25. This is the phosphatase that actually regulates the phosphorylation on CDK1. By inhibiting both of those together, you're actually able to dephosphorylate CDK1 actively and no longer phosphorylate it. We think that gives a really exciting synergy and an opportunity to further the efficacy beyond what we see with a single agent. You've done SNIPRx screens for PKMYT1 inhibition. What mutations have you identified as sensitive to PKMYT1 inhibition and what have you seen with RP-6306? Yeah. As you know, we did this for RP-3500 and for all of our late-stage programs as we do a step two chemical synthetic lethal screen, where we take cells plus and minus the inhibitor, knock out all the genes in the genome, and look for those that hypersensitize the cells with the genetic aberration to RP-6306. When we did that, we identified one gene in particular that we've talked about, which is called FBXW7. It's an E3 ubiquitin ligase that normally regulates the levels of certain proteins in the cell, including cyclin E1. When it's mutated in cancer, it actually results in the elevation of those proteins. that we've shown is mimicking the phenotype that we saw from cyclin E1 amplification and helping to expand out the patient population we could look at. Can you talk about the potential addressable patient populations with 6306, you know, looking at the CCNE1 amplification and the FBXW7 mutant? Absolutely. I think there's two key things to remember here, and that is that cyclin E1 and FBXW7 are mutually exclusive with one another outside of hypermutator tumors, where they probably don't really have a primary role, but a secondary role. They're unique patient populations. But the main ones that we're focused on, at least in the initial stages, are ovarian and gastroesophageal due to a very high level of cyclin E amplification in both of those diseases, as well as knowing that cyclin E is an initiating event in those tumors. FBXW7 is particularly high in colorectal cancer, where there's a huge unmet medical need being around 13%-14%. There's a number of other interesting areas outside of that, including things like squamous cell carcinoma of the lung, where it's about 7.2%. You're in phase I with RP-6306, and you expect data next year. What should we look for in the initial data? You know, will it mainly be safety, or might we see initial efficacy data? And you know, just what could that look like? We're really focused as this is a phase I, as you point out, unlike TRESR and 3500 where it was a phase 1/2 by design. As a phase I, we're really only focused on safety, tolerability, PK/PD, and our hope, our highest aspiration is to be able to provide any initial vectors we can on a recommended phase II dose and schedule. You'll recall the rigor with which Marie and her team pulled together that very Project Optimus friendly and rigorous sort of approach to finding dose and schedule. Multiple doses in a three-dimensional chess game. I think you should expect the same kind of rigor this time, and that's what we're guiding to. Okay. You're also conducting combination studies. Can you just talk about those studies and, you know, when we might see data from those? Yeah. We've launched three studies. The first was the GEM combination, which was launched in December. Subsequently, we launched a bit later into this year a Pol theta combination and a RP-6306 combination. I know that the RP-6306 combination, sorry, RP-3500 RP-6306. The combination was completely carved out of the Roche deal, so we kept that to ourselves. It was already planned and is now underway. We haven't given any specific guidance to when that combination data will be ready for prime time. It started, you know, the December launch of the gemcitabine combination was nine months after we started the initial study. You know, we're excited about the combinations. We wouldn't have launched as aggressively as we did into these. You know, as Mike alluded to earlier, you know, and as you can also refer to in that Nature paper, we saw really compelling tumor growth inhibition in all our preclinical work and monotherapy. We saw, you know, substantial and durable tumor regression with the combination. We're excited to be able to get that data out and, you know, we're just cautious because we're seeing a real slowdown globally in recruitment in oncology sites, which we've talked about before. Really a lot having to do with the Great Resignation, not anything specific having to clinical trial work, but it's been a concern and surprising since we didn't see anything like it in the first two years of COVID. Great. Well, in the last few minutes, maybe let's talk about RP-2119 as a Polθ inhibitor. You know, what makes Pol theta an attractive target, and how might RP-2119 be differentiated from other Polθ inhibitors? Polθ, maybe to step back, I think, the cell has three mechanisms for fixing double-strand breaks in the cell, three main mechanisms. Homologous recombination, which is defective in many tumors through loss of BRCA1, two, and other agents. Non-homologous end joining, which only works in certain stages of the cell cycle. It has a backup mechanism that's driven by Polθ, called microhomology-mediated end joining, MMEJ for less of a mouthful. When you see tumors that are mutated in HRD, they have this reliance upon POLQ to be able to maintain the integrity of the genome. It's been seen through a number of studies that were most recently presented by AstraZeneca and others at ASCO, where they've even shown that you can see this signature of POLQ activity as you're looking for regressions due to that agent. The real excitement is the ability to combine and to look at single agent in later stage disease, but also potentially to go into earliest stages of disease in combination and hopefully changing the evolution of that tumor by blocking the mechanism that's driving mutagenesis and allowing to get a lot longer durability in those patients in those early settings. Really those both of those key spaces are really, I think, extremely exciting and something that we're really pushing hard to get into the clinic and to be able to ask. We haven't revealed what component of Polθ we're targeting. But what we have said is there are two domains that everyone is focused on targeting, both the polymerase and helicase domain. We have made inhibitors to both. We've taken the whole of the data biologically as well as chemically to bring forward 2119, which we believe is the best approach for us to be able to answer all the questions we want to in the clinic, both from a drug perspective and from a biology perspective. We'll reveal more about that next year as we progress towards the clinic. Well, great. Looks like we'll have to leave it there. Thanks so much for your time. Thank you, Jeff Hung. Great to see you, and thanks for having us here.