Good morning, everyone. Tara Bancroft here. I'm one of the senior biotech analysts at TD Cowen. Thanks very much for joining TD Cowen's Fifth Annual Oncology Innovation Summit. For our next session, we have a fireside chat with Mersana, and it's my pleasure to introduce Marty Huber, who's the President and CEO, and Brian DeSchuytner, the COO and CFO. Marty and Brian, it's a privilege to have you both here, and thanks so much for joining me. Before I get started with questions, I wanna address those in the audience. Please feel free to submit a question to me via email or the link provided, if you have any questions that you want addressed. So I guess either Marty or Brian, can you start with a general overview of Mersana and a quick update before we get into specific questions? Thank you, Tara. I'll start on that one. It's just to kinda give a little background. Mersana is a company who focuses on ADCs. Our kind of core hypothesis is despite the recent advances in the ADC space. I mean, there's obviously multiple products on the market that are providing benefit for a lot of patients. We believe there's still an opportunity for better ADCs. All of the current ADCs have platform-limiting toxicities. These are things that are not related to the target, but just come with the platform and limit the dose and the ability to combine with other agents. Our core hypothesis is that we can create innovative new scaffold linker payloads that overcome these dose-limiting, off-target platform toxicities. When we look at where we are as a company, we have two products that are currently in the clinic, and ongoing. One, our lead is on our Dolasynthen platform, which is a cytotoxic anti-tubulin payload that's innovative, that we created. And this is targeting B7-H4. It's called XMT-1660. And then our other program is truly innovative is 'cause not only cytotoxic payloads, but it's a STING, you know, an immune agonist payload, and that's XMT-2056. That's a HER2-based antibody, and that's also early in ongoing dose escalation. Okay, great. So I guess let's start with the platform. So can you first start by discussing some of the learnings that you've gleaned from the Dolaflexin platform, and how you used that to improve and develop the Dolasynthen platform? One of the things we learned is while our payload, our Dolaflexin payload, had the opportunity to improve over MMAE by being more controlled in the tumor microenvironment, the first-generation polymer, the Fleximer polymer on Dolaflexin that we used, had a challenge. It was a natural product, so it's heterogeneous. So you had varying range of payload for each ADC. That heterogeneity had consequences that we didn't fully appreciate until we were in the clinic, and what it led to was a lack of charge balance for these payloads, which led to kind of a non-specific uptake in endothelial cells, which led to fatigue. We saw these grade three AST elevations. And then, unfortunately, we also saw thrombocytopenia, which was associated with a set of rare grade 5 bleeding events, which altogether are associated kinda with this non-specific endothelial injury. We now our second-generation platform was Dolasynthen, and we've learned preclinically and in data we shared most recently or earlier this year at ESGO and SGO, showed that Dolasynthen, by being a homogeneous DAR-6 with the appropriate charge balance, doesn't carry these same liabilities. So what we believe is what we've learned about Dolaflexin is our novel payload solved the problems of neutropenia, neuropathy and ocular tox that we saw with MMAE. It had these additional toxicities. Dolasynthen, we believe, solves for neutropenia, neuropathy and ocular tox, but also solves for our first-generation platform's liabilities. Okay. Yeah, that makes sense. So I mean, I know that you generally went over it, but maybe just a more detailed recap of this, the 1592 data that you presented at ESGO. You know, I think that this is a data set that was still largely underappreciated by the Street. So, like, in what ways did that help validate the improvements of Dolasynthen over Dolaflexin and show you more of what's target-driven versus platform-driven? Well, Brian, why don't you take that one? Yeah, well, so frankly, we're in a situation to really have the ideal clinical experiment. And we published this data at ESGO. We also published a follow-up poster at AACR, that's important as well. You know, this is the ideal experiment because 1592, which is our first Dolasynthen, has the same target, the same antibody, and the same payload as UpRi, right? Our Dolaflexin, otherwise known as 1536. So what's different between the two is the linker scaffold, right? Because UpRi is intrinsically heterogeneous, it had some of these liabilities that Marty described at these transient but maybe high-grade AST elevations, nausea, fatigue, thrombocytopenia, and these infrequent bleeding events. Because 1592 is homogeneous, what we observed was a substantial mitigation of those toxicities, the top toxicities of UpRi. You know, this is, you know, the tables presented at ESGO and AACR are cross-trial comparisons, but the differences are night and day. And, you know, we ascribe that to the homogeneity of the Dolaflexin, or the, excuse me, the Dolasynthen platform, and the resulting efficiency of delivery of payload to the target versus non-target tissues. What we did see on 1592 was dose-limiting pneumonitis, and we ascribe that to an on-target liability that we observed first with UpRi, but observed in an enhanced way with 1592 because of this efficiency of delivery of the payload to the target. In our AACR poster, there are some IHC studies that really call into stark contrast where in our rat toxicology models you see NaPi2b in the type two pneumocytes, a type of stem cell that is responsible for repair of the gas exchange surfaces, and in elderly humans living with cancer, can be in a proliferative state. We can also stain for where 1592 delivers payload, and it delivers it to those exact same cuboidal type two pneumocyte cells. You know, just side by side, it's very apparent that that is an on-target toxicity. So that provided us a lot of excitement about the potential of the platform to mitigate off-target platform-type toxicities. But it also helped us understand that the on-target liabilities of NaPi2b are considerable. Yeah. So then, considering that, can you then talk about what you really like about B7-H4 as a target, especially, in contrast to NaPi2b? Yeah. I mean, for us, B7-H4 is an exciting target that's gotten even more exciting as we look at some of the other data out there. First of all, as Brian mentioned, one of the challenges of NaPi2b, there was a proliferating cell type that was, you know, normal tissue, that expressed high levels of NaPi2b. With B7-H4, in our preclinical data and our explorations of human tissue, we haven't seen that same liability. In addition, we've seen from our competitors' data, no real obvious evidence of an on-target platform toxicity associated with B7-H4. So based on that, and then maybe, well, the final thing is, it's also expressed widely. Whereas NaPi2b had kind of been narrowed down to ovarian cancer as a primary target, NaPi2b, the B7-H4 is present on ovarian, endometrial, and in breast cancer, both hormone receptor-positive as well as triple-negative. So we think for multiple reasons, B7-H4 is a much better target for our Dolasynthen platform. Yeah, and you so you mentioned that others have validated this lack of on-target tox, but can you now discuss what advantages that 1660 has over other B7-H4s that are in development? I think, Brian, do you wanna? Yeah, sure. So, I mean, I think there are four assets, including ours, that have been in the clinic for a while. Pfizer is developing an asset based on the Seagen VC-MMAE platform, and it suffers from the same limitations as all of those agents, neutropenia and peripheral neuropathy. They showed about a 20% response rate in triple-negative breast cancer and ER-positive breast cancer at a recent investor day. And that was enriched for response in higher expressors, which is very pertinent data, right? It really confirms the bimodal nature of the expression of B7-H4. You're either very low or very high, and that's frankly ideal from the perspective of drawing a cutoff for a biomarker. But, you know, they have had to abandon the more intensive regimen that they were pursuing for toxicity issues. Hansoh and GSK have an asset that's delivering a topoisomerase I payload. They show, just like Seagen, responses in a number of different tumor types, which we think is great validation, but do show some myelosuppression that in some cases can be severe. I think the more significant issue for the topoisomerase delivering ADCs, which include Hansoh and AZ, is that probably the highest unmet need and the fastest market opportunity is settings in breast cancer, where patients have already seen Enhertu and Trodelvy. And probably in the U.S., 80% of patients will have seen one or both of those agents. And so, being a third in sequence topoisomerase inhibitor could be challenging in that, in that space. And there's quite a body of literature that's been emerging over the last several scientific conferences to suggest that the acquisition of topoisomerase resistance in sequential therapy, and there was some at ESMO, at SABCS, and there'll be some more at ASCO as well coming up. Okay, and. Well, in addition to the favorable tolerability profile that we see on Dolasynthen, the idea that we are not subject to some of these same resistance mechanisms, I think, are important elements for differentiation and the ability to combine in the future. So, you mentioned a lot of interesting things, but, specifically, you know, among all the preclinical data that you've gathered for 1660, do you have data that you guys have shared in the past for the ability for it to be sequenced post-topo, or anything else that you wanna highlight? Yeah, we continue to work on developing data to support for how it integrates into the current standard of care. But there's a very detailed and comprehensive paper on our website in Molecular Cancer Therapeutics that really gets into a good deal of depth about the preclinical work to support the selection of 1660 as a clinical candidate. That includes some of the DAR ranging work that we've done and work in different PDX and CDX models. Okay, great. Hopefully, hopefully people go and check that out. So, let's move next now to the ongoing trial. So can you first overview a little bit of the design on that, where you are in your dosing exploration and scheduling combinations? Like, what have you learned so far as the trial's progressed? Thank you. I mean, one of the things to step back on the design is this is a pretty straightforward dose escalation, three plus three, for the initial part of it. A couple of the things we did a little differently, though, was we did include these cohorts we talk about a lot about backfill, and what that allows us to do is, and this is our discretion with the investigators, is open these up at any given dose schedule combination, and more specifically, we can actually say in a given indication. So in the trial overall, we're enrolling patients with breast cancer, or either hormone receptor-positive or triple-negative, but also endometrial and ovarian cancer because these are the tumor types that have the highest expression for B7-H4. Importantly, we're enrolling anyone, regardless of biomarker expression, but we are gathering that data. So what we can do is we kinda put all four of those indications in during doing the dose escalation part. We do that to keep the speed going, to make sure you keep going up. But then importantly, we can drill down with these backfills and get a wider patient experience in a specific tumor type. One of the things we've learned, as Brian alluded to, in XMT-1592, we ran into this dose-limiting NaPi2b-associated toxicity at 56 mg/m². So even though that was on target, as we understood it, you still have a little bit of nervousness as a company when you're, you know, you've had a couple products where you've run into some liabilities. We were proceeding along, and if you think about it, though, as you start removing off-target, you know, on-target and only off-target, you get pure platform tox. And what we've learned with Dolasynthen, we've been able to go much higher than we originally anticipated. As we disclosed previously, we're at 59 milligrams per meter squared and continuing to escalate. So one of our learnings was, is if you get rid of the platform tox, you really can go higher, and, you know, fortunately, and unfortunately, from a guidance point of view, it's longer than we had originally guided, so that was a little bit of a painful learning, but overall, good news that we're still going up. The other thing we've been learning as we got to understand our molecule more was the other lever you wanna pull when you're trying to optimize the dose is also look at schedule. Are you trying to get a less frequent dosing with very high concentrations in Cmax? But as we look at how other ADCs work in our own data, it's more apparent that maybe it's less about how high can you get the dose, but how long can you sustain a critical concentration? And so one of the advantages we had by continuing dose escalation is that in parallel, we're starting to explore some alternative schedules. You can ask, "Why are you doing that?" One of the reasons is, is one way to do this is you go all the way up, see as high as you can go, and then if you then play with schedule. But we actually said, "You know what? If we're gonna continue to escalate beyond where we are, we know we're at clinically active doses. We, as we disclosed, we've seen responses." We thought now was a good time to start working in parallel on: Are there alternative schedules that will also optimize the dosing schedule? So... And then I think maybe the final learning for us was, you know, everybody's been talking about Project Optimus. I'm kind of an old-school guy who was one of the guys who are the reason probably that rule got written because I'd get somewhere near a dose, start it, do pivotal studies, get on the market, and then clean up dose problems later. I mean, how many... Brian and I worked together on Zejula. There's a drug that the best dosing schedule came well after we were on the market. Keytruda, we changed the dosing schedule. So one of the things that became kind of a reality as I talked to my peers and looked at what other ADCs were doing, is, this time the FDA was serious. And so one of the things that became very clear is the more data you can understand about your dose schedule, PK/PD, before you get into expansion and pivotal, really does potentially save you some time on the back end. If you look at some of the others out there, they're taking three and four doses into large phase II expansion sets. So while all this extra work we're doing doesn't necessarily mitigate the need to do a second dose in pivotal studies, you know, hopefully, we can, you know, take two doses into, as opposed to four doses into pivotal studies. So it's a long answer, I know, but there's, in this Project Optimus era, dose escalation and backfill has gotten a lot more important to really get a fulsome data set. Well, that makes sense. So it's entirely driven by the rationale for the complete exploration to satisfy the FDA and to optimize the dosing schedule appropriately, and is not driven by any safety signals that you've seen, right? Well, we're not going into the 1660 data, but I think, you know, what we've shared on 1592, we're confident in the Dolasynthen platform. And, you know, well, once again, we're not getting into the data. We feel. Mm-hmm. You know, we're pretty confident that the ILD we did see with the 1592 was on target, as Brian shared about with our AACR poster. Okay, great. So I guess let's- Which is also on the website. Which is also. Yeah. Great. Let's talk about the data that's coming up for 1660. So, I mean, in general, how should we think about expectations for what we'll see, not only in terms of the types of data points that you're actually going to provide, and potentially how many patients, but also for what you would consider to be good data? Yep. We're not giving any specific guidance on upcoming efficacy data, but I think there's public data out there that we can... You know, we always refer to is one is, for triple-negative breast cancer, and we're probably gonna focus there primarily for our conversation today, is standard of care chemotherapy is 5%, and that's at best. So this is an incredibly high unmet medical need. The number that most people are looking at, though, is the recent... You know, Brian referred to the Pfizer update. They had a 20% response rate in triple-negative breast cancer, as well as in hormone receptor-positive. One of the reasons we wanna be a little cautious in just jumping on that bar and saying, "That's the number to beat," is what was not in that data set or disclosed about that data set was how many of those patients had prior Trodelvy and prior Enhertu. We know their program includes ex-U.S. sites. That's even more amplified for the Hansoh program because that was China-only data, and we know that utilization of Trodelvy and Enhertu in general is much lower there, if at all, to be frank, in that patient population. So one of the challenges is your data representative of what the current triple-negative environment in the United States looks like, which is patients will have seen, or the vast majority of patients will have seen Trodelvy? One of the things we're gonna be very... We believe is very important is, given some of the resistance questions Brian raised earlier, is does. You know, our hypothesis is our agent should show evidence of meaningful anti-tumor activity in patients who've progressed on Trodelvy, for example. So we think that's important data to share. But when people start doing the comparisons, we may have shared that data, but it's not clear what will be the response rate in that population from the others. Mm-hmm. Yeah. And I, and I think just to follow up, one of the challenges with the GSK Hansoh data is that that was entirely enrolled in China for the data disclosure at ESMO 2023. So it's hard to answer that question about prior Trodelvy or Enhertu, because that might have not been an available therapy to those patients. Yeah. And so just to be clear on that, so when you do report the data, you do plan on providing different subgroups like that, like those that received prior topo, those who have not? Ahead of time, can you provide the U.S. versus ex-U.S. split for enrollment? We've only been operating in the U.S.. We are U.S. only, and while we can't—we don't have the exact number. You know, we don't share the exact number. The vast majority of TNBC will have seen at least Trodelvy and/or Enhertu at some point in time. The one other. So we will share the available information on that specific topic. Okay, great. And so I know we're quickly running out of time here, but I wanna give you the chance to go into a little bit more detail on Immunosynthen, and STING, and the 2056 program. So, you know, you already provided a very brief update, but, you know, can you talk about the ongoing phase I and what you are kind of expecting and hoping to see there? Thank you, Tara. First of all, what we'd like to remind you, it's a HER2-positive population, and people go, "Well, HER2 is very crowded. Why are you doing that?" That's, in a way, that's the best place for us to go. We know HER2 is a validated target. Our question is, "Can our Immunosynthen, bringing in a STING payload," so you have this completely novel approach of activating the innate immune system, "can it fundamentally provide benefit in patients who've actually seen prior HER2 therapy with traditional cytotoxic payloads?" So for us, this is a game-changer because it's the opportunity to show that you can take, you know, let's say, let's take breast cancer, gastric. These are generally not. Most of those patients have very cold tumors, and the ability to activate the innate immune system to make them warm or make them hot is would really be transformative, 'cause that'd be a completely new entry for ADCs. So for us, it's not only about, does that particular molecule, XMT-2056, have the potential? But we think that opens up, you know, a wide range of targets for application of our Immunosynthen platform. Okay, perfect. And so I know that you recently started, I guess restarted, enrollment. Did you make any changes to the protocol, and could you just describe currently what's going on with enrollment there? Well, the main thing we realized is that humans are much more sensitive to STING than primates, and murine, and rodents. So what we've done is we dramatically lowered the dose. It's a dose that we still believe has the potential for meaningful activation of the immune system, but at the same time, it's well below the doses that we originally started with. We're going into dose escalation. At this point in time, we're not... Other than continuing dose escalation this year, we're not giving any further guidance, and maybe that's a little bit of a lesson I learned on trying to give guidance on dose escalation, where you don't exactly know where your dose is going to go. So we'll be a little cautious on what we share on that. I think other than to say, it's up and running at sites. We have some excited investigators, and we look forward to generating some data. Great. Yeah, so do we. Okay, so I, you know, we just ran out of time, but before we close out, I really do want to ask you guys, what aspect of the Mersana story do you feel is most underappreciated by investors? Well, I think one of the critical issues is we recognize the value of 1660 and 2056 as assets, and we think people are looking at that. What I think may be underappreciated is, beyond those, is assuming we show strong validation of the platform concept, further validation with 1660 for Dolasynthen and, and 2056 for Immunosynthen, we believe that opportunity is there's a wide range of targets which could use a better scaffold linker payload. So we think once we kind of get these initial proof of concepts done with these two clinical assets, there's a tremendous opportunity for a very deep and extensive portfolio that we can build off of that. Great. Okay, well, you know, Marty, Brian, thank you both for being here, taking the time to educate us on Mersana, and thanks, everyone, for listening. Thank you. Thank you.