Morning. 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 Bicycle Therapeutics with CEO, Kevin Lee, CFO, Lee Kalinowski, CSO, Nick Keen, and CMO, Dominic Smethurst. Welcome. For those who may not be familiar with Bicycle, can you provide a brief introduction? Sure. Hi, everyone. I'm Kevin Lee, CEO of Bicycle, and thanks to Jeff and the Morgan Stanley team for the invitation. Bicycles are actually small peptides, constrained peptides, so very different than what you'll be probably used to. These are peptides, typically nine to 15 amino acids in length. What differentiates them and what gives them these structural constraints is a small molecular weight scaffold. We call it a scaffold, a small organic molecule that we put at the center of the peptide and creates three covalent links to create a bicyclic structure, hence the name Bicycle. We think this conveys upon the molecule's very differentiated properties. To list a few, maybe Nick might jump in and add to the list I'm about to share. Firstly, they're small, and that's really important for us because unlike antibodies, they can rapidly penetrate into tissue. The second thing that's important is they're fully synthetic, and that's really important when you think about, you know, manufacturing cost of goods, etc. The third thing is even though they're small, they have a large molecular footprint, and that's very good for drugging molecular targets that have been historically very difficult to drug. We have a phenomenal success rate at drugging a whole range of targets, which are classically intractable to small molecules. Then the fourth thing I'd mention is that they're very easy to conjugate. It's very easy to attach to a bicycle another molecular entity, whether it's a bicycle or a small molecule or whatever you care to mention. There are technical reasons for that, which probably don't have time to go into. We think bicycles have a suite of properties that are very differentiated from small molecules and from large molecules. We think there's an array of different therapeutic opportunities for them. Where we've chosen to focus our initial attention is oncology and the idea of using bicycles as precision guidance systems, because they can penetrate into tissue very easily, very rapidly, and because they can drug some of the targets that are, you know, typically associated with the biology of tumors, which are classically intractable to small molecules. We think bicycles are excellent at delivering a range of payloads into tissue. One of the things we've been doing, frankly, for the last two or three years is validating that hypothesis using toxins as the conjugates. We've chosen two oncology targets to explore. One is a target called Ephrin A2, which is actually a tumor antigen which no one else has successfully drugged to date. We had announced some data last week showing that we can indeed drug the target and in a safe way and an effective way. We are by drugging that target, we can shrink tumors. The second target, which, if you like, to be the opposite end of the spectrum, is Nectin-4, which many of you will know is a highly validated target for ADCs, with enfortumab vedotin or PADCEV in the clinic. Here what we've shown is, yes, we can drug this target. We generate, I think, really good efficacy. But importantly, as is the case with EphA2, we have a profile that's very differentiated, particularly around the safety. So I think we've shown that Bicycle as a platform is able to deliver payloads into tumor effectively against an array of targets. On that basis, we'll continue to build other programs. Great. Well, before talking about the pipeline programs, maybe let's spend a couple questions on Bicycles. You know, you touched upon how they're differentiated from small molecules and nanobodies, but if you can talk a little about, you know, what enables the ability for Bicycles to drug previously undruggable targets? There's a couple of things about the Bicycle technology, which is worth mentioning. The technology arose from the lab of Sir Greg Winter, and many of you will know Sir Greg as the architect of antibodies. He developed technologies to humanize antibodies. He developed phage display, and a large number of the world's commercial antibody products, licensed products can actually be traced back to Greg Winter technology. What Greg's breakthrough in Bicycle technology, he looked at the antibody, and in his mind, you know, there are 1,500 amino acids as part of that antibody, and probably only a handful of those amino acids are actually involved in target engagement. You know, the Fab domains are very much there to provide the structure to enable those CDRs to interact with targets. Something that people forget is the Fc portion of the antibody has polypharmacology. He thought, maybe we can do better than that and reduce the active binding site of an antibody to a very small fragment. He came up with the technology at Bicycles, but we use phage display again. We use phage display for the discovery of bicycles, and that gives us a phenomenal advantage. We have libraries of diverse molecules in the range of 10 to the 17, which is a number you just can't imagine. We have very large, very diverse libraries, very successful at screening and getting molecules. It's the large molecular footprint that gives us the advantage. Small molecules, because they're so small, when you think about enzyme inhibition, for example, the only way you can inhibit the enzyme is actually by binding the active site, and the active site of that enzyme classically is conserved amongst the family of enzymes. It's very hard to get selectivity. However, what gives the enzymes their different physiology is slightly outside of the active site, the shelf regions around the active site, which confirms the selectivity for substrate, and that's where we bind our bicycles. We can block the active site by binding to parts of the molecule which are differentiated from other enzymes. Nick, anything you'd like to add there? No. I mean, I think the only other thing I would add is, you know, back to your first question was, you know, One of the things that really attracted me to the company is that, you know, there are technical drug discovery problems which cannot be solved with small molecules or biologics. People commonly refer to antibodies as being much more selective than small molecules. In fact, they're not. You know, antibodies evolve to interact with multiple receptors in the immune system, and that actually leads to some pretty significant flaws for antibody-based technologies. Everybody thinks antibodies are super specific at the antigen binding end, which they are, but they're poorly selective at the non-antigen binding end. That, for example, in antibody-drug conjugates, can lead to pretty severe, skin and eye toxicities, and that's a problem that just intrinsically can't be solved. We believe we can solve for that. We have the same kind of binding potencies, the same kind of binding affinities as antibodies, but in a far more selective package that we can apply medicinal chemistry to and modify the properties of. Very neat technology. It solves for a lot of issues in small-molecule drug discovery, as well as in large-molecule drug discovery. You previewed a couple of your programs, but actually there are different formats of bicycles. Can you just briefly talk about each type and when are they best used? Sure. I mean, we, you know, there are different formats in the way that we can connect bicycles to other things. For example, in one application that probably is most focused on today is coupling bicycles to toxins to deliver toxins, specifically to tumors. You can also couple bicycles to other bicycles, and in immune oncology, we're doing that to take a bicycle that recognizes a tumor antigen and hook it up to a bicycle that recognizes key receptors in the immune system. The platform is incredibly versatile in that perspective. You can really click these together like Legos, if you like. Then within the platform itself, we can, for example, vary the length of each arm of the molecule, so we can generate sort of asymmetric bicycles that can form tertiary structure in one arm versus the other to bind to particular parts of a protein. Actually, because these are synthetic molecules and not something that's just provided by nature, we can really change the physical properties of those molecules. For example, to really extend half-life if we needed to do that in a specific application or to reduce interaction with the liver or with other organs that you might not want to interact with. There's a lot of versatility in this platform. It's a fascinating sort of hybrid of classic small molecule discovery and biologic discovery. Great. Well, let's move on to BT8009. Early this year, you reported interim phase I data with responses in urothelial cancer that appeared durable with limited safety signals at the 5 mg per square meter once-weekly dose. Can you just remind us what you saw? Yeah. I'll ask Dom to jump in, but I'll give you the headlines, and Dom can provide a bit more detail. It's very challenging and we're trying to or people are trying to compare what we're doing with 8009 to PADCEV, which is a licensed product. Do you compare the activity we've seen in a phase I dose escalation with the what's on the license? Or do you compare the data we presented as part of the phase I dose escalation with what they presented in their phase I escalation and you know and then there's phase II and so on. It's a bit like spaghetti trying to do the comparison. What I'll try and do here is compare what I think is the right comparison, which is the phase I escalation with the phase I escalation. When we did the escalation, this is, you know, virtually the first time a product of this nature has gone into man, so we were really learning as we go along and refining as we go along. In their phase I dose escalation, they had patients, I think about 40% of patients had one therapy, only one or more. We had patients who had two or three lines of therapy. We're dealing with naturally, you know, a sicker patient population. Their response rate was in the sort of 38%, I think it was, response rate. Ours was early small numbers, but 50%. Our safety profile, as you already referred to, is quite differentiated. Importantly, there are scientific reasons for this, we believe, and Nick's alluded to this. We've not seen skin toxicity. We believe that's a function of the Fc domain and the multiple biology of Fc receptors. Our molecule, if you do a panel of 5,000 receptors and you put our molecule on that panel, there's only one protein that lights up and that's in the case of BT8009 Nectin-4. If you did the same with an antibody, any antibody against its target, you would see multiple proteins lit up because of the Fc biology. We don't see the eye toxicity. In this first generation of Bicycle molecules, the linker toxin we've taken from an antibody. It's an unoptimized Bicycle unoptimized linker toxin. We are seeing some of the deficits from ADC technology, so we still see some neutropenia, which we believe is very manageable. We do see some low-grade neuropathy. Based on small numbers of patients, I don't think it's anywhere near the same as been reported for ADCs, but we do see low levels of neuropathy, which our physicians tell us is low grade and manageable. Dom, anything you'd like to add? Yeah. It really is differentiated. Although we're kind of so close to it, and again, I get the constant reassurance from our investigator teams that many of whom have got experience with the antibody-drug conjugates as well, and they do say this is different. What's been really pleasing is they told me back in beginning of the year that it was different with the neuropathy as well. Now unfortunately, the headline rates we got in our 5-mg group really didn't look that different at all. I think it's just gonna be the interplay of time and durability, both with respect to the efficacy and safety readouts. Of course, those two are interrelated as well because the more the patient can stay on drug, the more they can maintain their efficacy. It is increasingly looking like it's differentiated, and I think we're just gonna have to hit that nail with the data hammer several times to get people on the same side of belief that we're on. You mentioned, though, that some of the investigators mentioned that it's different, but what has the reaction been from the investigators and other clinicians that? Yeah. They're very excited. We've had to kind of contain their enthusiasm during the dose escalation, but we had, at that time, seen efficacy at all dose levels, and we still continue to have excited investigators at our new dose levels as well. Yeah, what's been really lovely is they've started to learn how to play with it in their hands, and they've stepped away from the constraints and the conceptualization that they originally came to it with, which was around that this is an antibody-drug conjugate, right? The neuropathy is just gonna get worse with time. Yeah, it's been nice knowing you, but we'll have to leave you. It's not. They've given them dose holidays, they've given them dose reductions, and they've really titrated down to a point where they can sustain responses while not. The belief going into the trial was that it was just an ever-narrowing tunnel. Once you got a patient with neuropathy, that's it. It was a kind of countdown timer, and they'd be off in 2, 3 months' time. We really have been able to break free from that sort of constrictive paradigm, and I'm looking forward to sharing their work with everyone potentially later on this year. Of the 4 responders, 1 progressed at about 3 months. At the time of the interim data, you had responses ongoing to like 6-8 months. Just any update on the other 3 responders? You have to wait and see. You knew I was gonna say that. Yeah. Had to try. You're planning to explore additional less frequent doses and additional doses for phase II. Any updates on those fronts and the current thinking for the phase II? I think I mean, the 5-milligram dose ticks all the boxes for us. It's definitely a go forward dose. I think it was important for us to explore other ways of giving the 5 milligrams, 'cause that's what we're doing in the exploration we're doing. Pre-clinically, we find if you give the numbers I'm gonna say now are illustrative just for understanding. If we give 5 milligrams in a preclinical model system, 5 milligrams per week generates the same efficacy as 10 milligrams every 2 weeks. It was important for us to explore that dosage schedule and also a kind of 2 out of 3 dosage schedule. We view that the totality of that data is almost the same dose, actually. We're exploring that because it's really important, again, to mention nobody has put a molecule like this into man before. We're learning about the ways to give the drug in an optimal manner. We also think that while 5 mg per week is very well-tolerated, very durable, great efficacy, et cetera, you know, there is a small advantage to patients, obviously, if it were two out of three or some other less demanding paradigm that we use. We're exploring that. We're really excited to, you know, to be approaching the end of the escalation phase where we can really, you know, address a lot of the questions that people have. But again, equally, we have to do this methodically, carefully. We think we have something very, very special, and we don't wanna damage that by doing something out of turn or too quickly and manage it in the wrong way. We're looking forward to giving an update towards the end of the year and, yeah, see what people think. On that update, what should we expect to see? Like, would it be updated data, or is it the design of the phase II? Like, can you just talk a bit more about? I think you could expect to see both. There'll most certainly be an update. We've got lots of data that we wanna share. Yeah, some thoughts about how we will where we go from here. Okay. Great. Well, maybe moving on to BT5528. Last week you reported updated phase I data. For those who may not have followed that update, can you just remind us what you saw? I think fifty-five twenty-eight is a great example of what we've been saying. We gave an update at the Triple Meeting last year, and we had some, you know, some toxicities, not anywhere in the same ballpark as ADC technologies in my mind's eye. We had some toxicities. A year later, with the investigators understanding a lot more about how to use these molecules, those toxicities have diminished. What we showed, and this is the full escalation data. We showed that the molecule is active in EphA2 expressing tumors. Again, just to pull everyone back for a second, there have been multiple attempts at targeting EphA2. There's been ADCs, there have been unarmed antibodies, there have been antibody-targeted nanoliposomes, and they've all failed quite dramatically actually in the case of the ADC MEDI-547. In the very first cohort of patients, five out of the six patients had an uncontrolled bleeding event. The trial was abandoned, and other people have seen the same phenomena with the ADCs. The first thing we had to do with this program was show that we didn't have that same liability, and I think we've proven that quite convincingly now. We showed that the molecule has activity in EphA2 expressing ovarian and bladder tumors to date, and we obviously will continue to explore other tumor types. We haven't prematurely instituted an arbitrary cutoff in terms of what we expect in terms of EphA2 expression. I think that proved to be the right thing because we are now seeing activity extremely low EphA2 levels. A tumor which doesn't express EphA2 doesn't appear to respond, but a tumor which expresses EphA2 does expect to respond. The numbers of patients we're talking about are... I know you people will obsess about the numbers and the percentages, et cetera, in a phase I dose escalation, but they're relatively not small numbers of patients, so I can give you the numbers. It was 2 out of 9, so about 22% of EphA2 ovarian cancer patients had a RECIST response. In one case, a complete response, in one case, a partial response. The complete response patient has been on therapy now for 17 cycles. Yeah. Yeah. 17 cycles. You know, 22%, 2 out of 9. If you know, go to the other extreme, 2 out of 3 bladder patients, 67% response rate, if you wanna get excited, responded with partial responses. We've shown the responses are durable. They are actually, they develop with time. Importantly, the safety profile is just spectacular in my mind. You know, we have 0 TIN events, grade 3s. Actually, I don't think we have many grade 1s. We have 0 grade 3 neuropathies. We have 0 ocular tox. The only across 45 patients or 15 patients at the go-forward dose, I should say, two cases of a transient anemia and one ocular. Was it not grade 3 ocular? I don't think it was grade three, though. We have some blurred vision. Yeah. Dry eye syndrome. One patient. The profile is very differentiated. We're beginning to see activity. The other point to make is, you know, these are very sick patients. I think the median lines of therapy was about 7 prior lines. Very pleased with the data, very pleased with the tolerability profile. That was the main thing for us in this trial. We're seeing quite nice responses. Dom, you should jump in. Yeah. I think that the next stage is more of the same really. It's very pleasing and reassuring and perhaps sometimes these days necessary to have efficacy in the dose escalation, and we've got that. And we'll expand accordingly looking at EphA2 as well, continuing to develop and refine ability to manage the adverse events. Just giving them, like, prophylactic antiemetics and things, giving them permission to come back in and change their meds if they're getting some GI side effects. All of those things have really helped cut down the adverse event rate quite significantly. And yeah, we continue to expand. We look at other tumor types as well. Tumor types that we have selected in the expansion essentially are ones that we know have a reasonably good chance of expressing a high degree of EphA2, but also they're tumor types that have historically responded to other MMAE-bearing, typically ADCs as well. So we're hoping that enrichment for both target staining and intrinsic predisposition to the response to MMAE will yield further results as well. Then as Kevin alluded to, having in the expansion phase patients that will have had fewer prior lines of therapy, I think will give us a more indicative response rate and help us decide what we do from there on in. Great. You've talked about the next steps for the study, but what should we look for in the update that is expected next year? I think you can, you know, we'll be giving an update on where, you know, the expansion, where we are in terms of that. The follow-through for the escalation. I think they're the main things to expect. Okay. Great. Well, maybe in the last couple minutes, looking out at 2023 and beyond, you're working on next gen BTCs. How are they different from current BTCs? Would you initially use them for lifecycle management or expanding into additional indications? This is a really interesting question for us because what I think we've shown over the last 2 or 3 years is we've validated the technology. We've shown, you know, we've shown in a couple of cases that we can deliver payload into tumor, and that's something that not every ADC can do or other format can do. We now have a technology which we're very confident can deliver payloads into the heart of the tumor. We've been working quite hard over the last 12 months on our, you know, our screening capability is just really very special. We're now beginning to develop a lineup of everybody's favorite tumor antigen, plus others, binders to them all. We have a range of effector molecules that we're working on. We have the ADC validated linker toxin, the Val-Cit-MMAE, which we've used for our first couple of products. We have our own proprietary systems that we're very excited by, where we think the MMAE release liability will be reduced. We have our CD137 agonist. We've already said previously that we have our NK agonist. When you start thinking about the opportunity, we have a lineup of tumor binders, a lineup of effector payloads. Obviously, we—that's a lot of different molecules. When we have to be guided by the biology of a given tumor and where it makes sense to deploy each one of those molecules. I think the opportunities are quite exciting for us. Nick, anything? Yeah, I mean, I think Kevin's put it very well. I mean, you know, got unique technology. We know it's clearly differentiated from antibody-based approaches. It allows us to do a lot of things that have not been done in the past. We're clearly working on, you know, just in the toxin delivery space, on different novel linker systems that are tailored to the technology. There's room there for new payloads, both small molecule and potentially in time radioisotopes as well, which the technology seems very well tailored to. You know, in the immuno-oncology space, I think people have really not been paying attention to what we're doing there. The intrinsic properties of the molecule are so different than can be provided by biologics, that we're able to interact with immune receptors and produce biology, which cannot be produced by biologic-based approaches. I'm pretty happy to say that we've published on this very recently around CD137. Yeah, watch this space because there are some really exciting approaches coming. Not that the current ones aren't. We're super happy with those. This is a, you know, this is a big concept company with more than just one product. I think people are really gonna pay attention to that. Great. Well, looks like we'll have to leave it there. Thank you so much for your time. Thanks, Jeff. Thank you. Thanks, everyone. Thank you. Thank you.