Kymera Therapeutics, Inc. (KYMR) Earnings Call Transcript & Summary

Welcome again to the 40th Annual JPMorgan Healthcare Conference. I'm Eric Joseph, senior biotech analyst with the firm. And our next presenting company is Kymera Therapeutics. It's my pleasure to welcome the company's CEO, Nello Mainolfi, to talk to us a little bit about the company. There is a Q&A session after the presentation. Feel free to submit questions by clicking the icon over there, and I'll work them in where appropriate. So with that, Nello, thanks for showing some of your time with us this morning.

Thanks, Eric. Thanks for the invitation to participate in the conference. I'm excited to share with everybody our story today. So first of all, maybe just to remind everybody, Kymera has been evolving quite rapidly in the past 2, 3 years. I would say, thanks to an amazing team of dedicated, I call it, R&D entrepreneurs, but also thanks to supporting both partners and investors. So today, I do want to talk about our pipeline progression and the exciting science that we're doing here at Kymera. But also I want to make sure it's clear where we're going in the next 3, 5, 10 years and why we believe Kymera has the potential to have a profound impact in a wide variety of diseases and indications and realize our vision to becoming a fully integrated global biotech company. So starting here for Slide 3 here, very simply, the most basic principles of why we believe we have such a profound opportunity is really -- and actually, I like to call the responsibility to advance one of the most transformative technologies in modern medicine is actually the fact that we have, even from the technical point of view, a unique opportunity. And this comes to really being able to finally capitalize on all the knowledge that we've created tying diseases to genes and then to proteins and now having finally a technology that is able to edit the human proteome. And this technology, as you see on the slide here is, in the moment, we believe, the only technology out there that is able to selectively, specifically and reversibly edit the human proteome. So targeted protein degradation starts with a very simple concept. And the concept is coopting, a well-established cellular machinery that is the Ubiquitin Proteasome System to drive degradation of disease proteins. And how we do that is using, as you see in magenta here on your slide, this heterobifunctional molecule. You can also call them bispecific molecules that binds specifically to a protein of interest, in this case, green protein that, let's say, we assume is a disease-causing protein and brings it to the Ubiquitin Proteasome System via the E3 Ligase, which is a protein in blue. And then basically, the job of the bispecific molecule is done and that the Ubiquitin Proteasome System recognized the target protein and destroys it through their -- the proteasomal core. So what are the clear advantages of this technology. So we don't need to inhibit the function of the protein. We just need to bind to the protein of interest, so expanding the potentially druggable proteome. It's highly specific and catalytic. So a small amount of drug can actually have an effect for a prolonged period of time. And it results really in a genetic like knockdown with the flexibility and the ease of development of small molecule development. And more importantly, this technology, for the first time, we have something that is both protein type and disease agnostic. So you can target any type of protein in human cells, in every tissue, in every organ and for every disease. So it's not surprising that we at Kymera and many others, I'm convinced, obviously, share the thought that this technology is well poised to expand the currently drugged proteome, which again, we're still driving about 15% to 20% of what we know is in human cells. And so we have a unique opportunity to expand that with this technology. So something about Kymera. So I think our vision that you see on the bottom left there is really the north star that has been guiding us since the founding of the company. I was fortunate enough to be there about 5.5 years ago now. And we continue to invest in both pipeline, platform, additional technologies, partnerships as well as human resources to build a disease- and technology-agnostic, fully integrated global biopharmaceutical company that uses both the concept and all the technologies that come with the concept of protein degradation to deliver a new generation of medicines. So where we are today, I think it's fair to say a recognized leader in protein degradation. We obviously want to grow to be a recognized leader in the biopharma space. We, as I mentioned, invested in building a fully integrated company. Initial internal focus has been so far in immunology, inflammation and oncology. But thanks to our partnerships and especially one that we have with Vertex, we're able to work in 5 other different diseases outside of those and makes Kymera uniquely a platform that is disease-agnostic already. Again, we've talked about partnerships. We have one with Vertex, one with Sanofi, a way to forward integrate the company and to continue to grow in a derisked manner where it makes sense, whether it's new diseases or it's going in large immunological indications. Last year was a big year for us, 2021. We established several firsts in both -- at Kymera, but also for protein degradation in general. We had the first placebo-controlled randomized Phase I study of a heterobifunctional degrader with our KT-474, where the first heterobifunctional degrader that is targeting an undrugged transcription factor with KT-333 that is now in the clinic as well as, hopefully, the first targeted therapy in diffuse large B-cell lymphoma with our KT-413 degrader. But that work doesn't stop there. We remain committed to develop at least one new clinical program per year. And so for this year, we have KT-253, which is our MDM2 degrader that is planning to go in IND-enabling study as is already in file and will clear in IND in 2022. We continue to be focused on innovation, and you'll hear me talking about advancements that we made in our platform as well as new investments that we made in that space. We're well capitalized. The cash position that you see there is based on September 30, but we're still in a really unique cash position for a company of our stage, of our size to invest both aggressively, but also thoughtfully as we continue to scale the company. So it's an upshot of our pipeline and then I'll try to go into highlighting each of the programs. So KT-474, we believe this molecule has potential to be best-in-class, small molecule, anti-inflammatory drug for a wide variety of immuno-inflammatory indications. And you'll see -- I'll highlight some data today that hopefully will substitute even further my statement. KT-413, so these are now the oncology pipeline that is now in clinical stage. KT-413, as I mentioned, potential to be targeting, for the first time, a well-defined genetic population with MYD88 mutant tumors in lymphomas, in Waldenströms, in CNS lymphoma, as you see in a few minutes. KT-333 is our first drug from our STAT3 franchise. This is targeting both liquid and solid tumors. We don't have time today to talk about our immunology and fibrotic efforts with this target, but I'm sure that we'll have opportunities to do so as the year goes on. And then our p53 stabilizer, I think our approach of degrading finally MDM2 and for the first time, really unlocking the full biology of p53 stabilization. It's a program that we're very excited about. I'll have an opportunity to highlight some of the data that we believe makes this molecule KT-253 a really unique drug in the p53 space. As I mentioned, we have several programs in the discovery stage. We expect to have at least one program from our tissue-restricted platform enter development later in the year. So starting from -- I might have skipped the slide. There you go. So before we go into both our platform and pipeline, I just wanted to share our philosophy around target selection. And there is a biological and clinical point of view, which -- I've said it many times, so I actually decided to remove that slide, which is we want to work in pathways where there is high degree of validation, both in terms of human genetics as well as human pharmacology. Basically, other nodes in those pathways have been drugged already, but where key nodes have not been drugged with any other technologies or not well drugged so that protein degradation will provide a unique unlocking solution. And we see true to that philosophy in every target that you see Kymera work on. So that's the really biological context. Now there is a technical context. And our commitment that we shared in the R&D Day in December has been about driving all target classes in human cells. Again, with that biological and clinical philosophy in mind. And so we've divided in a way, all target classes that we can think of in these 3 categories. There are ones that are inadequately drugged with clear degrader advantage. So basically, there are small molecule inhibitors, but those are insufficient at best to drug that target fully. Examples, IRAK4, examples, MDM2, heterobifunctional, highly specific protein degrader can unlock the new biological power in that mechanism. Then in the middle, you see the category that, to be honest, in the field of protein degradation in general, but I would say in the industry is general is at the lower return on investment, which is going after undrug -- some call undruggable targets. So basically, targets that no other technology can really go after. Transcription factors, adaptor proteins, obviously, all intracellular proteins, where either the technology is absent, or the delivery methods are absent. And so we've already made inroads. And again, the first company to go after one of those, STAT3, which is now in the clinic is an example of those. There are targets, though, that are nonligandable with small molecules. So the subcategory, we define it as undrug and unligandable. And this is where we've devised with some of our collaboration, A-Alpha Bio on the West Coast, University of Washington and NYU, a new strategy to unlock high-value, undruggable, unligandable targets through our molecular group center. And the goal there is to identify novel pairs on novel E3 ligases with novel targets of interest that have already inherent natural affinity that can be augmented by small molecules, creating novel surfaces that create productive protein-protein interactions. We have already some exciting early programs in this space, again, not working on cereblon, not working on IMiDs that we believe have the opportunity to change the landscape of undruggable, unligandable proteins out there. And then the third category, which I think is also a category that Kymera has been owning for the past few years is enabling clinically validated biology to be fully explored by using tissue restricted or selective E3 ligases that allow for drastically improved therapeutic indices that can now allow those mechanisms to be fully explored in the clinic, both as single agent and in combination. And you'll see an example in the next slides. So what I summarized here in our platform slide is our commitment to novel E3 ligases. We have several that obviously, we haven't disclosed yet, but are making inroads into our discovery pipeline and eventually become clinical candidates that we will be discussing about or at least part of clinical candidates. We are a company committed to understanding PK/PD. And I think the data from our Phase I study with KT-474 shine lights to how we think about translating preclinical into clinical. And then the depth as well as the breadth of data that we generate around those levels of understanding. Our chemistry platform is really focused on a wide variety of deliverables, but one and the foremost, I would say, is identifying and developing degraders with optimal pharmaceutical properties, and then, in parallel, having the right technologies to identify ligands for this undruggable, unliganded, so far proteins. And we use a comprehensive approach here that goes all the way from in silico, AI enabled to high-content screening like [indiscernible] libraries through covalent screening, et cetera. And then we've talked about our molecular glue efforts. So I will not spend more time on that. So an example of our tissue-restricted E3 ligase platform that comes alive into this slide, program, we have a target well-characterized, clinically validated, toxicity that limits the application, both a single agent and combination. Toxicity driven by one cell type in the bone marrow, and we're able to find -- and we've been able to find an E3 ligase with low or almost absent expression in that particular cell type. And in doing so, we've been able to build a degrader that degrades the target in cancer cells, as you see here on the top right, does not degrade the target in the blood cells. And again, that's obvious because E3 ligase is not present. And these results in activity in vivo where we're able to demonstrate that we have full viability of this particular blood cell with a degrader that, again, spares pharmacology in that particular blood cells. And if you instead use a small molecule inhibitor that inhibits this target more broadly, then you see a drastically reduced viability, which eventually results into tox. So this -- just to give you a sense of where our platform is and, finally, after having discussed some of these approaches that we've had for a while, finally, that some of these programs are entering development. So I think what I've shown you so far is our commitment to building a pipeline that is differentiated based on target selection, our commitment to develop drugs that are going to transform diseases respective of how challenging it is to develop those drugs, a commitment to building a fully integrated company. So that hopefully gives you an idea of Kymera's approach, which, I think, is differentiated at least. But maybe using this slide is an opportunity to go and maybe more specifically as some of the approaches that we have, which I think are quite unique. So we've talked about already the target selection. Our platform investments, both on the tissue selective E3 ligases as well as our new approach to identify novel generation of molecular glue in the same company that does heterobifunctional molecule, that does tissue-restricted E3 ligases is quite a unique approach here to own all these target classes in the human cells. I think the way and our philosophy on how we think about development of this drug, understanding PK/PD first and then going into disease is very important. We've talked about our TPD first, so I will not cover that again. And again, it's a serious commitment to innovation that hopefully comes through the slides as we go through the presentation. So I thought I'll spend the next few minutes here on the programs that we have, both in our clinical pipeline and then spend a couple of minutes also on MDM2. So IRAK4 is the program that is more advanced with KT-474. Very quickly, what is the hypothesis here for our investment? The [indiscernible] pathway is one of the most validated pathway in innate immunity, where we have cytokine blockers for the [ IL-1 ] family cytokines that signal through the [ IL-1 ] receptor. We have also blockers for cytokines that are secreted through these pathways like TNF, IL-6, IL-17, IL-23. So clearly, this is a central access for innate immunity. And the industry has been trying to find the small molecule solution to blocking this pathway [indiscernible] wide variety of diseases. The closest we've come is in IRAK4 selective small molecule inhibitor, which actually has continued to validate this approach with positive proof of concept in [indiscernible] from our colleagues at Pfizer, but we also learned that the kinase function is insufficient to block fully the signaling of TLR and IL-1R receptors. And in fact, there is another function of the IRAK4 protein, which actually enables the MYD88 complex from functioning. And the ability to remove the protein is able to block both the kinase and the scaffolding functions and allows us to fully block the TLR and IL-1R signaling. And for the first time, have a small molecule agent that is able to capitalize fully on the biology of this pathway. What also helps us is that we have, in one of those few cases where we have really strong human genetics, both on the biology, but also on safety that tells us that adult humans that have an IRAK4 [indiscernible] mutation are healthy individuals. So we've always felt confident that removing the target will not lead to any particular safety event. So as a summary of the premise here, our IRAK4 degrader has the potential, thanks to blocking fully the pathways, which is broad and well-tolerated anti-inflammatory effect for a wide variety of diseases. So not much time on this slide. This -- obviously, this will be on our website, but we look at the opportunity here almost in 2 categories: Th1-Th17 biology and Th2 biology. We believe that the IRAK4 and the L1 pathway in the TLR are relevant in both of these types of inflammatory processes. And so as we enter Phase II studies with our partners at Sanofi, we will be taking diseases from each of those categories, HS, AD, RA, potentially asthma, to continue to both validate and advance this therapy in all those patient populations, for which there is still a huge unmet need, especially in terms of small molecule oral daily drug. So a brief summary of the data on the left, highly selective degrader. IRAK4 is really the only protein that KT-474 downregulates. On the right, if you take human cells and you activate them with both the TLR and IL-1 agonist in order to suppress the signal, you have to degrade the protein. You see in orange. This is the only way to block IL-6 production that cannot be done with a small molecule inhibitor. Our Phase I study, so we've actually just announced yesterday, we've completed the dose escalation in the healthy volunteer portion. So we've completed the Part 1 and Part 2 SAD and MAD. And now -- and we've recruited more than 100 subjects to our study in less than a year, actually last year. And now we're analyzing data to move into the Part C or Part 3, which will involve AD and HS patients that will be dosed daily for 15 days with our drug to confirm the PK/PD that we've seen previously. So really highlights of the data. This is, again, PK/PD, namely, this is the PD portion of our MAD study. This is going from 25 to 200 mg. You see from day 1 to day 14 profound impact on IRAK4. All of those doses suppress the target more than 90%, with the top 3 doses reaching lower limit of quantitation. And then after we stopped dosing at day [ 15 ], you see a nice recovery eventually to baseline. This is a snapshot, just to show you that the day 7 and day 14, we reached roughly a steady state, and you'll see that we're able to suppress the target fully with maximum 98% degradation. In the skin, which is both the surrogate tissue, but it's also a disease relevant tissue, we started with very low level of IRAK4. So we're able to, even with very low level, suppressed target expression profoundly with dose-responsive manner, reaching almost lower limit of quantitation. Something to point out is that at day [ 14 ], we haven't quite reached steady state. So it's likely, although, we don't know yet, that continuing dosing will drive even further degradation, even at lower doses. And on the right, it just shows you that regardless of the starting level of IRAK4, we're able to suppress and degrade the target to the same endpoint, which is reaching lower limit of quantitation, which makes us feel very good about the kinetics and how this will play out in many different cell types and diseases as we go into heterogeneous populations. So just to summarize where we are, given the time, we've completed dose escalation. We've demonstrated both proof of mechanism and proof of biology in both SAD at higher doses and then [ MAD ] at much lower doses, I've only shown you the MAD data, and we've been able to show a full suppression of target in both blood and skin and a really profound effect on cytokines up to 9 different disease-relevant cytokines with maximum inhibition at about 85% at 100 mg dose. Blinded safety shows that the compound is safe and well tolerated and with really no serious adverse events. Upcoming milestones is initiating our open-label HS and AD patient cohort, sharing data later in the year, both proof of mechanism and proof of biology in those patients, and then with our partners in Sanofi initiating Phase II studies. So going to the IRAKIMiD very quickly. Why did we build a dual or maybe more technically right, a triple degrader? We are very obviously well versed in the MYD88 biology. And we wanted to build a targeted therapy in lymphoma selecting for MYD88 mutation. What we found was that IRAK4 selective degradation is insufficient to drive single-agent activity in MYD88 mutant tumors. What we found that there are contiguous pathways that almost start as compensatory pathways. And one of those was the interferon pathway influenced by Ikaros, Aiolos, IRAK4. And so we found that by degrading both IRAK4 and Ikaros and Aiolos, with the same molecule, so eliciting both the heterobifunctional degradation, but also molecular glue degradation mechanism. We're able to have single-agent profound antitumor effect. And so the indications that we're prioritizing, but by no means this would be the only indication that we're going after. But for now, our MYD88 mutant cancers in diffuse large B-cell lymphoma, Waldenström and primary CNS lymphoma, in which we have presence of MYD88 mutation that goes all the way from 25% all the way to 90% or more. So a real opportunity here for a real unmet need in this patient population. So this drive behaves really well, both in vitro and in vivo. We've always been able to show that [indiscernible], is more potent than IRAK4 alone and IMiD alone. So here, we're taking CC-220, which is a late generation IMiD that we're able to show both in vitro and in vivo to be much more potent. What's also important to highlight is that while we believe that we have the direct path to registration as single agent. As you know, in many tumor types, combination is something that eventually you will find to drive even more responses and then better benefits to patients. And so here, we found that our drug combines really, really well with active drugs like BTK inhibitors, BC-2 inhibitors or Rituxan. And you see suboptimal doses of KT-413 are able to show responses that are durable for months. And so this bodes really well for our strategy to enter also combination trials, potentially in earlier line of therapies. So just to conclude here, the IRAKIMiD part. I think we've shown that this drug has potential to really be the first precision medicine in lymphoma. And our goal is to develop both as in monotherapy in MYD88, diffuse large B-cell lymphoma as well as most from -- and primary CNS lymphoma, and then, in combination, ideally in earlier line of therapies. Just briefly, a couple of minutes on STAT3. So obviously, this is a larger franchise opportunity. We have cell intrinsic mechanisms, but also, let's call it, tumor cell extrinsic mechanisms, where we have an immunomodulatory effect that actually can be used both as a therapeutic angle in oncology as well as in inflammation. So some of the opportunities that we have here clinically are focused on some T-cell lymphoma and leukemias that you see here on this slide, where we know there is high percentage of either STAT3 mutation or activation in the pathway where we know that these tumor types respond exceptionally well to our degraders. And then we have opportunities as we started to uncover in preclinical studies of combining with PD-1 antibodies because we have seen that our immunomodulatory effect actually sensitizes tumor types to PD-1, especially the ones that are not highly responder to PD-1 therapy. So just quickly, you see in these lymphomas and leukemias, we need actually only 2 or 3 doses either once a week or once every 2 weeks to elicit full tumor regressions that are very durable even after we stop dosing. So this is clearly the type of data that we obviously would love to see in the clinic, and that we hope that will allow us to get a fast track towards registration for this drug as single agent. In solid tumor, what we've seen, this is a CT-26 model. This is KTX-201 now. This is a 2 compound. It's not our clinical compound, KT-333. We see that synergizes with anti-PD-1, and you see it leads to about 60% complete responses in this CT-26 model. We see that these complete responders reject tumor rechallenge. So obviously showing that there is immune memory in these tumors. And also you see that on top, the combination, but not the single agents, are able to extend survivals quite drastically. So again, we've told you STAT3. This is first-in-class agent for this target in the clinic opportunity, a single agent for liquid tumors and potential opportunities as combo in solid tumors. So to wrap up with MDM2, as I mentioned, the opportunity is very vast. I think we've been working on this pathway as an industry for many, many years. P53 is the largest tumor suppressor out there. And if you look at the DepMap graph below, these are all a large panel of tumor types, both liquid and solid. And you see that almost all p53 wild-type tumors are sensitive to the absence of MDM2. And so this is important because the absence of MDM2 confers sensitivity, the inhibition of MDM2 does not, at least not as much. And this is really the reason why we've started this program. [indiscernible] more than 50% of tumors are potentially sensitive to this mechanism. So small -- why small molecules don't work if you inhibit MDM2-p53 interaction. There is a transcriptional regulation that leads to more MDM2. And so given these small molecule are based on occupancy, and if you generate more MDM2, you really cannot keep up with the stoichiometry of inhibiting something that continues to be upregulated. So you end up with low activity and tox. By removing MDM2, we obviously circumvent the feedback loop. We drive apoptosis that happens very, very quickly. And so we see much more potency, and we've seen, in our preclinical studies, also an increased therapeutic index. So just to give you some snapshots of key data, you can see here on this slide, on the left, degradation of MDM2. So this is a picomolar degrader, 0.4 nanomolar. And then you see an inhibitor increases as we just said, MDM2 levels. So if you -- as I said, the correlation of p53 stabilization is with the absence of MDM2. And so you see in the middle, the way that a degrader stabilizes p53 is many, many fold, many, many orders of magnitude better than a small molecule inhibitor. And this results with potency at killing cells, tumor cells, that is at least 200-fold than the best small molecule MDM2 inhibitor out there. And again, just to reiterate, if you use a degrader, you remove the protein, and you clearly don't have MDM2 anymore. If you give an inhibitor in the bottom, you see an increase of MDM2 expression. And so in vivo, what happens? We see actually a single dose of our degrader is able to drive full regressions that can last several weeks. 1 mg per kg gives about 3 weeks of full tumor regressions, 10 mg per kg can actually last for almost 60 days. And you see on the right our commitment to apoptosis through very quick biomarker [indiscernible] p21, PUMA, which are well-characterized p53-related apoptotic biomarkers. So the indications that we're planning to go are large and potentially vast. And I think, for now, we just talked about AML, uveal melanoma and lymphoma, but there are others that we're exploring, where we're expecting to see really profound apoptotic response. And those are the ones that we will prioritize. So the molecule will enter the clinic towards the end of the year, early next year, IND filing this year. So what we expect in 2022, and then I'll wrap up, completion of our Phase I cohort with 474 and transition with Sanofi -- to Sanofi, proof of mechanism for the 2 oncology program, 413 and 333, IND filing for KT-253, the first tissue restricted E3 ligase in development as well as additional programs in both oncology and immunology reaching development as well as continue to maintain leadership in the space through investments, publications and collaborations. So I'll stop here, and happy to take any questions in the last few minutes.

Thanks for that presentation, Nello. Very comprehensive. Little time for questions here. There are a couple that have come through the portal. So maybe I'll select from those. One is related to the platform and asking you to sort of describe the tools in hand to find natural ligands for undruggable proteins and how you can ultimately make them more selective. And, at the R&D Day, you talked about an initiative to develop molecular glues with collaborators. Is there a meaningful difference in whether using heterobifunctionals versus molecular glues are best suited to pursue undruggable -- to the undruggable proteins?

Yes. No, I'll try and answer both of them very quickly. So the first one is, so our approach to find small molecule ligands to undrugged proteins are, again, technology agnostic. We use in silico approaches, fragment screens, we use [indiscernible] library screens, we use covalent screens, we use [indiscernible]. And I think what actually helps us the most is that we have a large investment in x-ray crystallography, in cryo-EM and in silico approaches to really assess ligandability of proteins. And when there is a high priority protein, we, most cases, throw all of these technology at it. There are a class of proteins that we consider ligandable, where all those technologies, we will yield ligands. There are classes of proteins -- I'll throw some like NRF2 or potentially [indiscernible] that have surfaces that have not been amenable to small molecule binding. So for the ones that we didn't find a small molecule ligand, heterobifunctional degraders, for us, are always the go-to technology because we control, we understand and we can drive specificity in the most straightforward manner. But for the ones that you cannot find a discrete binding event, that's where the molecular glue concept comes in. So this is not because we wanted to do molecular glue. This is because there is a need there. It's not a large number of protein needs, but the ones that are at there are really high value. And so what we're doing there is looking for E3 ligases that have natural affinity for some of these proteins and then augment that with increasing the specificity of the surface using small molecules. So that's how we think about the 2 spaces.

Okay. Okay. Great. Unfortunately, I think we're on to leave it there for time. But I want to thank you again for joining us at our conference, and thanks, everybody, for tuning into the presentation.

Thank you.