ADC Therapeutics SA (ADCT) Earnings Call Transcript & Summary

Good morning, good afternoon. I'm Eugenia Litz, Investor Relations at ADC Therapeutics. It is my pleasure to welcome you to our webcast today, highlighting our ADC platform and solid tumor pipeline. I would like to remind you that this presentation may contain forward-looking statements, which are subject to certain risks and uncertainties. Please refer to our SEC filings for more information. On our call today is Chris Martin, our CEO; Patrick van Berkel, our Head of R&D; and Joe Camardo, our Chief Medical Officer. And with that, it is my pleasure to turn the call over to Chris.

Thank you, Eugenia, and welcome to the ADCT pipeline -- solid tumor pipeline call. It was almost exactly 20 years ago that I sat down and met John Hartley in University College London to look at some pyrrolobenzodiazepines data in a couple of solid tumor models, in cisplatin resistant ovarian and glioma. These PBDs have been designed specifically to be active in hard to treat tumors, tumors which are resistant and refractory. And the data was truly startling. And that was the genesis of the formation of a company called Spirogen. Ten years later, we span ADC Therapeutics out of Spirogen to develop a PBD ADC pipeline. And now 10 years on from that, we have 1 drug approved, Zynlonta, which has been marketed in the U.S.; a second program, Cami, which is well on its way to a BLA filing. And behind that, an exciting pipeline of 5 solid tumor programs. And it's those that we're going to focus on today. This last decade has been extremely productive. As you can see here, we have these 7 programs, 5 in the clinic, 2 in IND-enabling studies. And behind that, multiple programs in research. And as I said, today, we're going to focus on these solid tumor program. And for the first time, we'll be talking about ADCT-212, targeting PSMA, a highly optimized and differentiated program that Patrick will be able to describe to you in some detail. So this pipeline has been produced by a world-class R&D organization, with all of the capabilities you need to go from target, right through to approved drug and beyond, as demonstrated with Zynlonta. It's a very capable program, a team based in London for research. They've recently moved to this state-of-the-art facility that you can see in these pictures, at I-HUB in the West of London. And the clinical team that's based in New Jersey and Lausanne, that takes these clinical candidates and turns them into oncology drugs. So it's my pleasure today to introduce you to Patrick van Berkel. He was employee #1 at ADC Therapeutics. He's the architect of the full pipeline, and he will be taking you through the solid tumor pipelines asset-by-asset. And also to Joe Camardo, our Chief Medical Officer, who's been responsible for the development of these drugs and is taking the solid tumor programs forward. So Patrick, it's a great pleasure to introduce you. Patrick?

Thanks very much for this introduction. And thank you all for listening in. I'm really pleased to update you on our research and development activities as well as our solid tumor pipeline. But first, let us take a step back. So what really is an antibody drug conjugate as we develop it? An antibody drug conjugate is an antibody that targets to a tumor-specific antigen conjugated to a toxin. Now we all know toxins, very known is chemotherapy, and they have been very useful for the treatment of cancer patients. And we all know that this typically is a mixture of different chemos and provide good efficacy. But the efficacy is very close to the toxicity profile of this chemo, which means that the therapeutic index of such a treatment is quite small, quite narrow. So the idea of combining an antibody with chemo or toxin is very smart, and that happened decades ago. And indeed, we've seen over the last 10 years or so that about a dozen ADCs have been approved in different indications. And indeed, give a real good benefit to cancer patients because they have a better therapeutic index. But of course, we're not there yet. We try to improve on the therapeutic index of our ADCs all the time. And we try to achieve that by improving the efficacy or reducing the off-target toxicity of our antibody drug conjugates. So the way we approach that at ADC Therapeutics is basically in 3 different buckets. Of course, the toxin. So the chemo as you call it, or the toxin, is really important. And Chris already mentioned the PBD, and we're proud that we have PBD chemistry in the toolbox. And later, I will explain what makes a PBD so unique. But going forward, we also think that it's important to have other toxins in our toolbox to develop antibody drug conjugates. And I'll explain in a minute also why we think that's critical. And in order to do so, we have in-licensed a very important technology for us, that's called the SiLent technology, and I will explain today what that technology brings to ADC Therapeutics. You can also think about a different concept. Instead of targeting a toxin to the drug -- to the tumor, you can also target, for instance, immunostimulant payloads. So these are not toxins on its own, but they really activate locally the immune system, such that the immune system will act the tumor more efficiently. And even you can think about combinations of things. You can think about combinations of different chemos, different toxins or a toxin with an immunostimulant payload. And that will, in the future, be quite important. But of course, in order to do so, you have to be able to conjugate all these different things to the antibody, so the tumor-specific antibody. And therefore, we have a strong focus on our conjugation chemistry such that it's site specific, but also that will allow us to conjugate different toxins, different payloads to the same molecule. Cancer treatment has been a treatment based on different modalities quite often. And therefore, it will be quite neat if you can combine different mode of actions in one product, and that's really what we think the future will go. And having this optionality in the toolbox is quite important for us. And finally, of course, antibody and protein engineering. Mean the antibody has to be tumor-specific. And how do you make it tumor-specific? Of course, it will depend on the antigen, but also you can make the antibody may be more tumor specific. And we approach that by different techniques such as masking or conditional binding techniques such that the antibody will have a higher preferential binding to the antigen in the tumor versus the antigen on healthy tissue. Within ADC Therapeutics, we also have what we call XDCs. We're not going into too much detail today, but we also have quite a few licenses for the PBD chemistry for what we call XDCs. And this means that it's not an antibody that you used to target, but you can think of a peptide or another protein scaffold that you make specific for a tumor antigen and you use that to conjugate the toxin to. The advantage typically is that these molecules are smaller and may be able to penetrate the tumor more efficiently. But of course, there's one thing missing on this previous slide, which is the tumor target. The tumor target is quite important and maybe the most important thing when you select and going to develop an ADC for a certain indication. The tumor target may be expressed on liquid tumors and may be expressed on solid tumors or maybe on both. There are even targets that are expressed in the both, on liquid and solid tumors. The expression of the target on the tumor may be heterogeneous, maybe homogeneous, can be high or low. So there are different, different parameters that you have to look in to. But typically, a target is not really tumor specific. There is healthy tissue expression, which can be high, it can be low, it can be on a critical organ or it can be on a less critical organ, but this is really an important component of the design of an ADC because this ties into how you select your toxin. Depending on all these parameters, you may want to go for a PBD, you may want to go for a PBD which is lower potency or you may want to go for a PBD of high potency because you think there is no risk for healthy tissue toxicity. But you may also decide that based on the tox profile or maybe based on the healthy tissue expression profile of the target, you think a PBD may not be the right toxin of choice, and you have to go for a toxin with another mode of action. So all these things come into play. And then, of course, the real hard work starts, and that is finding the right candidate for this particular target of interest. And that's a lot of work, I can tell, where you have to look into different parameters. So the linker is important, the toxin is important. You have different flavors of both. And of course, the same is true for the antibody. So you have many moving parts that finally have to give you the best candidate for your target of interest. But let's take a deep dive into PBDs. Chris already alluded to the PBD chemistry and how it was invented in the old days in London, and we're really proud that we have succeeded as a young company to bring the first PBD-based ADC to market with Zynlonta. That was quite an achievement of our team and credits for the entire team who made that happen. But let us explain how PBD works. So the antibody drug conjugate is targeted to the tumor body antibody is processed, and that processing delivers the toxin to the nucleus. In our case, a pyrrolobenzodiazepines. And the way the PBDs are developed is that they are able to cross-link DNA in the minor groove very, very efficiently. So you see the cartoon in the middle where the cross link is shown. So basically, once the PBD dimer has entered the nucleus, it will go into the minor groove and cross-link the DNA on any GATC sequence. So it's a GG cross-link. And these cross-links can be pretty persistent. And that was, of course, one of the reasons behind developing the platform, they had to be active in patients that were relapsed or refractory to standard treatments. So these PBDs cross-links are very, very resistant. So what does that mean? They sit and wait until the cell wants to divide. And basically, the cell can't divide because the cross-links will keep the 2 DNA strands together, and this will drive the tumor cell into apoptosis by causing a stalled replication fork, as you can see in the cartoon there in the middle. So the 2 strands of DNA are kept together and it can't separate, no translation will happen. So these cross-links fit very nicely in the minor groove of DNA, which means nothing hardly sticks out. There is no distortion of the helix structure. And therefore, the DNA repair mechanisms really have a hard time detecting these cross-links and that explains why they are so persistent. PBDs are very potent. So you can tell from the graph on the bottom right, that if you compare the potency of a PBD dimer to toxins that are more often used in ADC development such as auristatins, calicheamicins or maytansinoids, it's an order of magnitude higher than those toxins and a couple of orders of magnitudes higher than classical chemo. But the beauty about PBDs is also that we can tweak the potency of the PBD dimer. So now we're working also with PBD dimers that have a potency which approaches that of the old-fashioned toxins. And I will explain later why this is important for our product development. So just to summarize the benefits of PBD. So they were originally designed to maintain activity in relapsed patients or patients that were refractory to treatment. They cause DNA cross-linking in such a way that the DNA helix is not distorted, not disrupted, and therefore, the DNA repair mechanisms will have a hard time detecting these cross-links and they are, therefore, pretty persistent. They're very potent. We're allowed to work with what we call drug-to-antibody ratios of 2, which is quite technical, but basically, it means that we don't need a lot of toxin to add to each antibody, and that has some advantages in our drug development. And also, there is bicenter activity of the PBD. So this means that not all the cells express the target of interest in a tumor. So sometimes the expression is heterogenous. Now if a cell doesn't express the target, it can still be killed by a PBD that's released from a neighboring cell that has processed the ADC because it expressed the target. But that toxin then is released from the cell that's dying and is able to penetrate the neighboring cell and kill that cell. So in case of solid tumors, having bicenter activity we believe, is quite an important feature. And PBDs also cause immunogenic cell death, which means that once the cell is going into apoptosis, because of a PBD cross-link, it will cause a more immunogenic response against that cell, which is, of course, an advantage if you think about antitumor treatment. But also, we have to think, of course, about a couple of things with PBDs related to its mode of action. We must pick targets that are relatively clean, that are not expressed on healthy tissue to a high extent or at least don't harm critical organs. And of course, we also have to take into care when we do clinical development that this persistent cross-link may show up in patients. So when you look into dose escalation or dose finding, you have to take care of that delayed cross-linking. So let's take a step -- let's take a closer look at the PBD chemistry itself. So this slide shows you a structure of a PBD payload, as we call it, which means all the components that contribute to the PBD toxin linker. And how we're optimizing it to make -- to come forward with PBD-based ADCs that have an improved therapeutic index because we think we reduce the off-target toxicity. And therefore, we should be able to increase the dose levels of these PBD-based ADCs in patients. On the left you see conjugation. And why is conjugation important, because that determines how you conjugate the payload to the antibody. And of course, you want a stable chemistry to do that. So we use classical maleimide on the right spots and the antibody or we use things like bioorthogonal click chemistry to make it very stable. Another important component is the spacer as indicated here. We have shown that, for instance, if you change the [ pack ] spacer to something called hydraspace, you improve the off-target toxicity of the ADC at least preclinically, quite significantly. And therefore, in a few of our candidates, we have chosen to work with hydraspace as a spacer in the entire payload. The linker, of course, is important, whether it's cleavable or noncleavable. We believe that for solid tumors, a peptide-based cleavable linker is quite important because that will give you the bystander kill, as I just explained, a noncleavable linker will not give you that. So typically, for solid tumor programs, we will focus on cleavable linker. But last but not least, of course, we play and tweak the potency of the PBD dimer. And this example here is between a very highly potent PBD dimer called SG3199 and a less potent PBD dimer called SG2000. And later on, I will explain how that will impact the tolerability of the PBD. So you can see by playing with these different structures and different modalities of the PBD payload, we come forward with novel PBD payloads that will offset toxicity and hopefully will allow us to dose higher in patients. And this work has led us to a portfolio of 11 products or 11 product candidates. We have 11 target exclusive licenses for the PBD chemistry. And these -- the projects are shown here. There are some undisclosed, but you can see how they are divided between solid tumors, hematological tumors. And we also make a division between clinically validated targets as well as novel ADC targets. And as you all know, we've been strongly focusing so far on our hem portfolio in the bottom left with CD19, CD25 and CD22. And we're moving up now to the solid tumors, and that's what we are talking about today, in fact. So some metrics with this portfolio, we -- in order to get to this portfolio, we've looked at more than 170 different targets, whether we believe they were suitable as PBD ADC conjugates. We prototyped more than 35 of them in our facilities in London, and that has led to this matrix of products between the fourth quadrant you see there. We filed 8 INDs until to date. Our fastest time line to IND was 13 months. So it shows that we, as a company, can deliver quite quickly a product to the clinic, to be tested in the clinic. And once we are in the clinic, we have a very strong objective to go for a rapid clinical proof of concept just to see and to figure out whether we believe there is a way forward for the PBD-based product or not. And we do that by performing quite robust large Phase I clinical trials, as we've done with Cami and Zynlonta. And I think all of this shows and the success we had with the hematological programs that this is a very, very solid and robust approach, and we're therefore confident that this will also apply to our solar tumor pipeline, which we're going to talk about in a second. Finally, before we go there, I want to explain a little bit more detail about our Silinol SiLent linker technology. And why do we think this is important? Well, as I mentioned, there are a lot of targets out there, and some of these targets, they may be better off with another toxin versus a PBD. And therefore, to have a toolbox that allows us to work with different toxins is very important to be able to target any target we like. So we have in-licensed this new linker technology, which will allow us to basically work with every toxin that is out there. We've given here the examples of exatecan and MMAF, but these are just examples. We can stick basically every toxin to this particular linker system. Another important feature is the new novel novelty of the Silinol linker system. This is a linker system that is designed such that you can tweak the sensitivity to different pHs. And we believe that may be a very important, nice attribute of the linker system. For instance, the tumor is more acidic than healthy tissue. So we can design the linker such that the warhead is preferentially released at the tumor and not the healthy tissue, which has a higher pH. And finally, there is a nice feature of this system -- of this linker system, which is called the adapter. And the adapter will allow us to branch off from this one payload. So we can design payloads that have 2, 3, 4 toxins, for instance, per payload, but you can also think about combinations of toxins that you can make in one particular payload. I explained in the beginning that combinations can become very important. And this linker system will allow us to combine 2 different toxins or for instance, combine a toxin with an immunostimulant quite efficiently. So going on into deeper detail in our solid tumor pipeline. Chris already mentioned that we're going to talk about 5 different programs today, 3 are in the clinic and 2 are in the IND-enabling stage with ADCT-701 and ADCT-212. So the first time you're going to hear about this very exciting program targeting PSMA. But to start with Cami. So Cami, as you all know, is very advanced in the hematological setting. We just have completed the Phase II pivotal trial for Cami in relapsed/refractory Hodgkin lymphoma. But we also see a future for Cami in solid tumors. And I will explain in a minute what the mode of action or the mechanism is behind this thinking. It's a completely novel immuno-oncology approach. And we're going to show later that we have really encouraging data in patient biopsies that we took from the monotherapy dose-escalation study. So we think that Cami has a future in tumors that have been infiltrated with CD25-positive Tregs. And there's a long list of tumors that has been shown to be infiltrated with T cells and the long list is over there. You can read it. But it's quite a big list of tumors that -- where we believe that Cami may have a future. So where we are right now is we are currently in a dose-escalation phase of Cami with pembro, and Joe will later tell you about the next steps for this program. But first, let me explain why we believe Cami works in solid tumors because typically solid tumors do not express CD25. In fact, they never express CD25 and it's all about tumor-infiltrating T cells. So this cartoon here shows you that there are basically 2 types of infiltrating T cells. You have Tregs on the one hand and you have the affected T cells on the other hand. Now what you really need is a high level of proliferating effector T cells because these are the cytolytic T cells that will attack the tumor. The Tregs on the other hand are the immunosuppressive cells. So they dampen basically the immune response against the tumor. So what you want to do is change this balance and favor the affected T cells over the Tregs. Now the Tregs are CD25 positive, and that's where we come in with Cami. So the idea is that we deplete the CD25-positive Tregs in such a way that we increase the proliferation of the T effector cells, which will then kill the tumor. And indeed, in our preclinical work, this turned out to be the case. So here, we have pretty nice preclinical data that indeed demonstrates this mode of action. So what we have done before we could go there is to make a surrogate for Cami because Cami doesn't cross-react with mouse CD25. And of course, these are all mouse preclinical models. So what we had to do is make a surrogate, which we call Sur301, which basically is pretty similar to Cami, it targets CD25. It has the same payload, the same drug-to-antibody ratio as Cami. So it's a very nice surrogate for our chemi molecule. But we had to do that in order to be able to run these syngeneic models. And syngeneic models mean that these mice are fully immunocompetent and that's what you need because, of course, you're looking into the immune system of the mice. But what it does and what we show here is that indeed, in the syngeneic mouse models Sur301 is very active in a CD25 negative MC38 model. So on the left, you see that a single dose of Cami basically eradicates all the tumor. Those are the curves at the bottom. A single dose of 0.5 or 1 mg per kg was sufficient to get rid of all the tumor cells. You can compare that to PD-1 there. That's the blue -- the dark blue curve that doesn't do a lot, but there's some activity, but at least you see that Cami by itself is much more powerful than the PD-1 control here. But if we combine Cami with PD-1 -- Sur301, sorry, with PD-1 as shown on the right, you get very nice synergy. So in this case, we didn't take the high dose of Sur301. We took a suboptimal dose of 0.1 mg per kg, which on its own is hardly doing anything, combine it with PD-1, which on its own is hardly doing anything, and you get the light blue curves there shown on the right. So very nice synergy if you combine Sur301 with PD-1. We also showed that these animals build up immunological memory. So in the middle, you see that if we took the survivors of these studies, and we grafted these mice in the other flank with the same tumor cell line, the tumor won't grow anymore. And this is just because the immune system has built up immunological memory, recognizes the tumor and gets rid of it quite efficiently. And this happens after treatment with both Sur301 on its own or the combination of Sur301 and PD-1. Now of course, the big question is, indeed, is it explained by the knockdown of Tregs. And this is a busy slide, but I will take you through it just so that you understand what we're showing here. So here we are showing immuno-profiling of these mice during the course of the study from day 2 to day 11. And at the top row, you see the Tregs. So we counted the Tregs in the tumors. And where the arrow is pointing you see that already at day 2, there's a very strong depletion of Treg counts is mice treated with Sur301 or PD-1 and the combination with PD-1. In the middle row, you see that this is accompanied by an increase in CD8 effector cells. So where the arrow is pointing at 11, so it takes a bit longer for the effector cells to come up. But there you see a very strong increase in the CD8 effector cells, again, in both groups, the Sur301 alone or the combination of Sur301 with PD-1. And of course, as a result at the bottom, the ratio of T effector cells to Treg has changed dramatically. If you see from day 2 to day 8, you see an increase in this ratio again in both groups. So this shows that preclinically indeed, the efficacy of Sur301 is explained by depletion of CD25-positive Tregs and an increase in CD8. So we're happy to tell you also that this is also what's happening in the clinic. So in our Phase I monotherapy dose-escalation trial, we were able to acquire biopsies from patients that were dosed with Cami. And this slide shows you that in those patients where we do see T cell infiltration in the tumor, there is indeed a change in the ratio between effector cells and Tregs in those paired biopsies. If you compare pre-dose with the biopsy taken after cycle 1. And this happens both in the tumor as well as in the surrounding tissue. So basically, this confirms our preclinical observation that if you dose Cami in solid tumor patients, you're able to change the dynamics of the T cell infiltrate in tumor environment. We also looked at the blood. So this slide shows you data from ASCO 2021 and where we looked at Tregs and the T effector cells in the blood itself. And on the left, you see all the data that we obtained from patients that were dosed with 125 microgram per kilogram. As a course of time, you see that there's an increase in the ratio in that patient population, in that patient group. And you also see by the graphs on the right that indeed there is a correlation between the effector cells -- the ratio between T effector cells, Tregs -- and Tregs itself. So in the middle, it's a correlation plot of all the data we obtained from all these patients in that particular study. And we looked at the correlation between the change in ratio in T effectors with Treg, over Tregs itself in the middle or CD8 cells on the right. And you see that in the middle, this ratio is accompanied by a decrease in Tregs. And on the right, you see that at the same time, there's an increase of the CD8 cells, which is what you were hoping. So again, both the pad biopsy data as well as this immuno-profiling data and blood support, the fact that we're playing with the important T effector to Treg ratio in patients dosed with Cami. Now finally, before I hand over to Joe, I want to tell you about some other potentials for Cami. We already talked about the nice opportunity we see with PD-1 with pembro or the PD-1s, but we also have nice preclinical data to show that we see more potential for Cami, with radiotherapy and with IL2 therapy. So on the right, you see combination with IL2 therapy. And basically, it's the same model as I just explained for the PD-1 combination. So we dosed mice with IL2 alone, with Sur301 alone or with the combination of the two. And you see that either drug on its own doesn't do a lot. There's some mild activity. But if you combine the two, you get the curve in blue. So again, it shows there's a very nice additive to synergistic effect of combining IL2 with Sur301. On the left, you see the combination of Cami with radiotherapy -- sorry, Sur301 with radiotherapy. So here, we combined radiotherapy with a single dose of Sur301. And you can tell from the bottom right that this combination is highly effective. So in the combination, 8 out of 10 mice were tumor-free, while only 1 out of 10 with Sur301 alone or 0 out of 10 with radiotherapy alone. So it shows that combining these 2 treatments is very, very efficient in the tumor eradication. So for the future, after the PD-1 combinations, we see a lot of potential for Cami in solid tumors. So I'm happy now to hand over to Joe, who will give you some updates about our clinical progress of the Cami.

Thank you, Patrick. It's really inspiring to listen to you present the data from the lab, and not just for me but it's a motivation for the clinical team to get these ADCs and make them into cancer medicine. So thank you.

Sure.

Thank you. And with Cami development program, we started out with a strong foundation, a biologic rationale backed by data, not just in the lab, but as you see data in patients from their biopsies. So I want to tell you 3 things here: what we did, what we're doing and where we're going to go. On this slide, you see Phase Ia and Phase Ib. Looking on the left, from Phase Ia, we completed dose escalation in 44 patients. We have a dose range now, and we understand the safety and tolerability. There's 2 things about the safety. One is we see and understand the PBD side effects. And we're not seeing any surprises. Secondly, we don't see any Guillain-Barré syndrome in these patients, and that's important. So we are able to move with this dose range into Phase Id, that's what we're doing now. Using that dose range that we've established that as a single agent, which is standard for cancer treatment, we're moving up in the dose range with the standard dose of pembrolizumab. We have acceptable safety and tolerability now, and we have dose expansion proceeding according to the protocol, which means that we are allowed to have small expansions if we see activity or we want some additional safety data where we see a tumor that looks promising for our future development. We've already initiated a dose expansion at 60 micrograms per kilogram, but we're continuing to do the dose escalation, which is currently at 80 micrograms per kilogram. So dose range and safety from Phase Ia and as a single agent, ongoing dose escalation in combination with pembrolizumab. What's next? What we want to do is enrich the population for patients who are likely to respond to Cami and pembro together based on the mechanism of Cami. And we also want to fill a gap because there are patients who either relapse or fail to respond to pembrolizumab. And it's a totally different mechanism. I want you to look on the left side, where we will stratify patients based on Treg infiltration levels. Now why is this important? It's important because the CTLs, the cytotoxic lymphocytes that kill the tumor and the T regulatory cells tend to track together, and that's what that line shows you. So if you're measuring CTLs, you can be sure that there's also Tregs. We want to eliminate the Tregs. So that correlation is very important. And what we're going to be able to do is take a look at our data when we have responses and see how they correlate with the infiltration of cytotoxic lymphocytes and Tregs in the tumor biopsies. So we're using basic important data to guide the treatment and make sure that we're treating the patients who are likely to respond to this mechanism of action. And on the right side, when we move forward into our dose expansion, we're going to be using another measurable characteristic of patients' tumors, which is microsatellite instability related to DNA mismatch repair, which is dysfunctional, which seems to be a reason why patients relapse after PD-1, PD-L1 therapy. And that reason is tied to the likelihood that the Tregs infiltration is what's changing that balance for the patient. And what you see in the graph is that the elevation of cytotoxic lymphocytes occurs with MSI, microsatellite instability. So if we measure microsatellite instability, we can be pretty sure that along with the cytotoxic lymphocytes infiltrating the patient's tumor, there are Tregs infiltrating the patient's tumor. And that's the reason why PDL is not as effective anymore. So summary, we have a strong biologic rationale for camidanlumab combined with pembrolizumab. We have good Phase I data in a dose range with safety. We're ongoing dose escalation with again, good safety and tolerability. And we have, what I would call, a modern way of designating and identifying patients who are likely to respond to the combination and really fill a gap where pembrolizumab turns out to be ineffective. So with that, I want to turn it back to you, Patrick.

Yes. Thank you, Joe. So the next program we're going to talk about is ADCT-601. I'm not going to pronounce the INN name because I can't, to be honest. So ADCT-601 targets AXL. And AXL is a pretty well-known cancer antigen associated with poor prognosis, resistance to chemo and tumor immune escape. And I will explain in a minute what that means for us. We have very strong in vitro data and a very strong preclinical in vivo data in various models. Again, I will give you some examples. And AXL, of course, is known to be overexpressed in quite a wide range of tumors, including sarcoma, lung, breast, prostate, pancreatic, glioma and esophageal cancer. So where we are now is that we're about to start our Phase Ib combination study, and Joe will later tell you all about that design. But first, let's talk about AXL itself. So AXL is quite unique because it is expressed on 2 different compartments in the tumor. So there's a lot of data showing that AXL is also expressed on certain immune cells. We spoke about CD25 expression on Tregs. AXL is somewhat similar, but now AXL expression happens on what we call the M2 macrophages. And these macrophages have a similar role, they are immunosuppressive, which means that if you have a high infiltration of M2 macrophages into the tumor, it dampens the immune response to the tumor itself. So of course, if you're able to deplete those M2 macrophages, that, of course, would be great because then you would hopefully strengthen the immune response against the tumor. But AXL, in this case, is also expressed on the tumor itself. In fact, it's expressed on a lot of mechanisms that are involved in tumor escape and tumor resistance. So AXL, for instance, is expressed quite highly in metastases. And also, it's known that AXL is starting to overexpress upon resistance to standard chemo or other targeted therapies like EGFR target therapy. So here we have a protein that is expressed on immunosuppressive cells. It's overexpressed in tumor escape. And as you can tell from the immunohistochemistry analysis at the bottom quite often, you see tumors where there's AXL on both compartments. So it would be great if you can target AXL quite efficiently and deplete both the tumor as well as the immunosuppressive cells. So we've come forward with ADCT-601, which is our ADC targeting AXL. And here, we've incorporated the hydraspace technology. I already mentioned in the beginning that this is an important component of the PBD payload. And for this particular molecule, we've shown that inclusion of hydraspace offsets the off-target toxicity quite significantly. In our preclinical models, it was about threefold better compared to a payload that didn't have hydraspace in its structure. And this data was recently published by Francesco [indiscernible] in molecular cancer therapy. So it shows that by playing with the structure of the PBD payload, you can come forward with PBD payloads that have a better therapeutic index. Of course, they're still very active. So this data shows you that ADCT-601 is highly active in the pancreatic PDX model. And in this particular experiment, we compared the activity of 601 to an auristatin-based ADC. So you have the 2 curves where the 2 boxes are. At the bottom, you see the efficacy of ADCT-601 at a single dose of 0.3 mg per kg, quite active. But if you would apply that same dose of the ADC with an auristatin as shown in the other box, you don't see a lot of activity, basically is inactive. Only if you increase the dose of that auristatin-based ADC tenfold, you get similar activity compared to our single dose at 0.3 mg per kg with ADCT-601. So it shows it's a very active ADC in this preclinical setting. We also are doing a lot of work on finding combinations for our PBD-based ADCs and gemcitabine is one of those combination partners that typically works quite well with a PBD-based ADC. And that also turns out the case for 601. So this slide here shows you some in vitro data where we looked at additive or synergistic effects of gemcitabine with 601. And you see that it's always additive or synergistic. And also in vivo, we see a very nice combination treatment effect. So in the middle, you see the tumor volumes for the monotherapy treatments in green and red. And on gray, you see the combination of the 2, gemcitabine and 601 in this particular model. And you can tell from the right that the impact on the survival, if you combine the 2, it's huge. We go from all animals that are surviving at the end of the study. If you combine the 2, till about 1 animal-only surviving in the groups where only 1 of the 2 drugs was given. So it shows that combining the 2 seems like a smart idea. Of course, we're also looking a lot into indications where AXL is expressed. And this slide shows you some data of AXL expression in sarcoma. We spend a lot of time finding the right assay to detect AXL consistently in tumor samples and especially to detect AXL in the membrane of these tumor samples. And we now have a very nice assay that allows us to measure AXL in patient biopsies very robustly. And we're quite pleased to see sarcoma as one of our top indications where AXL expression is always typically quite high and on the membrane of the tumor, which is quite important if you think about the mode of action of our antibody drug conjugate. So I'm happy now to let Joe update you on our clinical progress.

Okay. Thanks, Patrick. In this case, we're beginning our clinical program with a couple of really important pieces of information from Patrick's laboratory. One is that you can identify tumors that express AXL that's really important, and that have genetic activity that's related to AXL expression. That's very important. So it will help us identify the tumor. Second, we have an indication here, sarcoma, which is very, very difficult and very resistant to current treatment. So that's an area of really compelling medical need. The third thing is Patrick showed you a slide where in this case, the potency of the PBD, highly potent, is very important. So we have 3 pieces of information to start our clinical program. So I'm going to tell you 2 things here, what we've done and what we plan to do later in the year. So what have we done? Phase Ia dose escalation study, not a surprise, very typical for cancer programs. We had 18 patients dosed. We established a dose range, which if you know anything about Zynlonta, you'll see it's pretty much like the dose range we studied in Zynlonta. So we're seeing, again, we know about the PBD side effects. We're not seeing any surprises. That's important. We have a dose range to work with. 15 patients had at least one post-treatment evaluation. Some patients had stable disease. We had 1 patient with sarcoma, who received several doses and actually had a partial response, which is consistent with what Patrick showed you about the expression of AXL in sarcoma. So not a surprise that we will be pursuing this in the future, okay? So we've reached the maximum dose, which is good, and we have an acceptable safety profile. The 2 things that you really need to identify in Phase Ia. And later in the year, we'll be moving on to the Phase Ib combination. And this reminds me to tell you the next thing that we learned from Patrick's lab, which is the combination with gemcitabine is synergistic. So in addition to the potency of the PBD, when we add the gemcitabine, we actually get additional effective response on tumors, which is really important. So here's where we are now in our stage for starting Phase Ib. On the left side, what you see is the dose escalation. Very standard except for the fact that we're using flat dosing, which means we're not adjusting by body weight. So we will go in ascending order. If you look at the top part of that left side, what you see is the ascending doses of 601 combined with pretty much standard dose of gemcitabine. Again, we're going to be looking at safety and tolerability. Down at the bottom, you see Arm B, that's dose escalation with 601 on its own without gemcitabine. And the reason we're doing these 2 separate dose escalations is because we have 2 separate cohorts that we want to test. One showing you on the top right is patients with sarcoma. As I said, sarcoma are highly resistant to cancer treatment, highly resistant. This is an area of compelling medical need. We'll study around 18 patients with a dose that we believe will be effective based on preclinical data and tolerability, and we'll be combining that with gemcitabine based on the preclinical data. So that will proceed once we get the recommended dose. And on the bottom, we're taking advantage of what we know about AXL expression and AXL amplification to study tumors in which AXL is shown to be highly expressed and with the genetic activity supports that expression. So those are tumors that are more likely to respond to the AXL-based antibody and PBD. So in summary, safety and tolerability, very good from Phase I single agent. We have a good plan for Phase Ib, which is a dose escalation, taking advantage of the known expression of AXL in sarcoma and taking advantage of the known synergy between the PBD and the gemcitabine, and we'll be starting this off later in this year. So Patrick?

Yes, Joe. Thanks.

Okay.

So now we're going to talk about ADCT-901, targeting KAAG1. So KAAG1 is a truly novel cancer antigen. I will explain in a second, some more information about KAAG1, gives you some more information about KAAG1. So we really see this as a very nice first-in-class opportunity for our company. So KAAG1 expressed in what we believe expressed in platinum-resistant ovarian cancer, fallopian tube cancer prostate, cholangiocarcinoma, renal cell carcinoma and triple-negative breast. So we just started the dose escalation of KAAG1 at the end of last year, and Joe will update you on our next plans. So more about KAAG1. We think KAAG1 has a very attractive expression profile. If you compare healthy tissue or normal tissue versus cancer. On the right, you can see tumor TMA, showing a lot of ovarian tumors as well as a lot of triple negative breast. And you can see from the brown staining, if you detect this with a KAAG1-specific antibody, that there are a lot of patients that are positive for KAAG1 in their tumor. In the middle, you see the healthy tissue expression. And although there is some expression in cervix and fallopian tube and a little bit in kidney, it's quite good if you look at all the other tissues, if not negative. And it's also important to mention that KAAG1 is, by definition, an intracellular target. So in healthy tissue, it's intracellular and somehow it's exposed on the membrane of tumor tissue. So that's another important discriminator between expression of KAAG1 in healthy tissue and tumor tissue. Now this protein was initially identified in the cDNA library of renal carcinoma cell line as an antigenic peptide recognized by cytolytic T cells. This was really back in 1999. Alethia is a Canadian company and had its substraction cloning analysis between healthy tissue and ovarian cancer tissue. And they had identified KAAG1 as one of the most abundantly over-expressed proteins in ovarian cancer. And they moved on and confirmed that messenger RNA for KAAG1 was indeed expressed in a high percentage of ovarian cancer patients as well as triple-negative breast cancer patients, and they confirm this by immunohistochemistry analysis of these and this was further evidence that KAAG1 is highly overexpressed in these tumors. There is some functional analysis, which I want to explain today. But finally, we acquired the portfolio, IP portfolio around KAAG1, from Alethia, including the antibody that binds to KAAG1, 304, which we used in the development of our clinical candidate, ADCT-901. And that candidate is shown here. So we have the antibody 304, that binds to KAAG1. And you see that the payload we're using here is tesirine. So the clinically validated molecule that we also use for Zynlonta and Cami. So no surprises with this particular PBD dimer. In the preclinical models, KAAG1 is overexpressed quite often, and we picked 4 ovarian cancer models, as shown here with different levels of KAAG1 expression. And we dosed the mice with a single dose of 901, and we also dosed the mice with a nonbinding control ADC called B12. This is a control we typically do. What you want to see is that the nonbinding control is basically inactive and of course, our ADCT-901 should be active. And again, this slide tells you that, indeed, in all 4 ovarian PDX models, we see very nice efficacy with 901 after a single dose of treatment, especially if you compare that to the purple curves, which is the nonbinding control, which is basically inactive in most of the models. So a very highly efficient ADC in ovarian cancer, and I'm looking forward to hear from Joe what the next steps are.

Great. Thank you, Patrick. Well, you shouldn't be surprised to hear that we're going into clinical development with a strong foundation from laboratory research, identifying KAAG1 as a surface protein that appears to distinguish normal tissue from cancer cells because of the expression. And the second thing helps to identify the various cancer cells that are actually expressing KAAG1. But probably the most exciting thing about this is it's totally novel and it's like a frontier for us that remains undiscovered. And we have a lot of opportunity here for something that's totally new. And this is really -- this is a great motivation for everybody in clinical. We're starting out with a standard dose escalation, not a surprise, 40 patients. However, we do have the ability to identify tumors that are likely to express KAAG1 based on the preclinical data. That includes cholangiocarcinoma, ovarian tumors, prostate cancer, renal cell cancer and triple-negative breast cancer. And some of these continue to be resistant to current treatment. So again, a medical need is very important. There's a long list of key selection criteria, but just 2 ECOG status is important in early clinical studies. And these are patients who are refractory or intolerant of existing therapies, which just a reminder that our clinical development programs depend on patients who are willing to be in clinical trials and who are in a state where there's really no other treatments. That's another motivation for us to do what we do. In this case, we're just starting. We started at a dose of 15 micrograms per kilogram. We moved on to 30 micrograms per kilogram. That's where we are now. We'll continue this dose escalation to establish a dose range, likely get to a maximum tolerated dose. Again, we're using a PBD, so we understand the side effects already. And when we see where we are with the dose, we'll be expanding using the dose that we choose from Phase I into a monotherapy program looking at these various types of tumors that are expressing KAAG and down the road, we'll be able to probably be able to measure KAAG expression and try to identify patients who are more likely to respond. Back to you, Patrick.

Thank you, Joe.

Okay.

So we spoke about CD25 Cami. We spoke about 601 targeting AXL, and we spoke about 901 targeting KAAG1. The last 2 projects are our IND-enabling programs. And the first one is ADCT-701 targeting DLK-1. Again, DLK-1 is a novel cancer antigen expressed in many tumors. I will give you a more -- I will give you more insight in that later. So we are going after rare neuroendocrine malignancies, such as small cell lung cancer, adrenocortical carcinoma, pheochromocytoma, paraganglioma and neuroblastoma. So as I mentioned, we are currently finishing the IND-enabling work. And the next step would be to work with NCI together on the Phase Ia study in solid tumors. So DLK1 and the expression profile of the DLK1 is shown in this slide. Here, we compare the expression of DLK1 in healthy tissue versus cancer. Now DLK1 is highly expressed in the embryonic stage and also in cells that have sort of stem cell like properties. But during the progression to the adult stage, that DLK1 expression is winding down. There are some cells that do show some expression of DLK1, but typically, it's quite limited, if not absent. So from that perspective, it's a very attractive target for ADC development. If you look at cancer, there's quite a long list of cancers that are reported to express DLK1. We have verified it in-house, and we think that a real robust expression of DLK1 is in certain neuroendocrine tumors. And therefore, we think that, that's the best way forward for this project, at least to start in clinical development. Of course, we can always go into other indications when we see fit. So 701, like 601, incorporates GlycoConnect technology and the hydro space. So basically, this payload here called PL1601 is the exact same payload as we used in the ADCT-601 program because we believe that offsets the off-target toxicity significantly if you use this PBD payload for this antibody. I forgot to mention, by the way, that we licensed the HydraSpace and GlycoConnect technology from Synaffix in the Netherlands. In terms of preclinical data, we are working with John Maris from CHOPS in Philadelphia. And they have a lot of preclinical models regarding neuroblastoma. So this slide here shows you 7 different neuroblastoma PDX models with different levels of DLK1 expression. In fact, one is negative. That's the one in the middle, at the bottom in the middle. And indeed, you see that in this particular model, ADCT-701 doesn't give any antitumor activity, which is explained by the lack of the DLK1 expression. In all the auto models, all the other 6 models, you see that DLK1 is present, at least for most of them, you can see there's one immunohistochemistry picture missing. But in the majority, there is the DLK1 expression at a high level. And this is accompanied by a very strong anti-tumor response of 701, the red curve. Again, you can compare it to the nonbinding control ADC in green. So again, it shows that 701 is a very active molecule in this particular setting in neuroblastoma. And this data has led us to go into a collaboration with NCI. Why NCI? Well, we're going into a rare disease area, neuroendocrine tumors. And NCI brings together all the leading experts in this space will help us to identify the patients and to get the patients into a study and also provide financial sponsorship for this particular study with 701, while we can maintain governance flexibility if we see that -- if we want to. So the next step here is to continue to IND and then start the dose escalation in this NCI-sponsored Phase I study, enrolling adult patients with all the indications I just mentioned in the beginning, all neuroendocrine tumors. If we see a signal, we will expand in a given indication where we do see efficacy. And secondly, we will expand in the basket with more indications at that dose -- at that particular dose. And finally, I'm happy to talk about ADCT-212. So ADCT-212 targets PSMA. And this is really a very nicely optimized second-generation PBD-based ADC targeting PSMA. So those of you who are following us for quite a while, we had a PSMA-based ADC with AstraZeneca and the clinic. And I will tell you later how we've changed from that particular molecule to ADCT-212. So PSMA is a well-validated cancer target in prostate cancer, but there also may be other applications of this ADC. PSMA is also known to be expressed quite highly in the neo-vasculature of a lot of solid tumors. So you could potentially also target these solid tumors, target the neo-vasculature, cause disruption of the vasculature and create antitumor activity. So we are currently completing the IND-enabling work for PSMA 212 and of course, hope to plan, submit the IND and start the clinical study later. So I mentioned that we had another PSMA-based ADC in the clinic together with AstraZeneca . And that one was called MEDI3726. The good news was that in the dose-escalation study, we saw clinical activity. However, patients couldn't really tolerate a lot of cycles, especially not at the high dose levels. And also, we found that PK of this ADC was quite rapid. And of course, if the PK is rapid, the exposure in the patient will be quite low. On top of that very fast PK, there was an instability. So basically, the ADC was instable and that further didn't help finding -- or getting the right exposure in the patient. So we worked around all of these issues and come forward with ADCT-212. First of all, we looked into changing the antibody in such a way that the new ADC has a much better PK profile than the old ADC. And this was done by changing the antibody to a fully human antibody. Secondly, we changed the PBD payload. So we used SG3249 in the previous candidate, and we're changing to PL1801, which is based on the lower potency PBD dimer, SG2000. And we think by doing so, we can increase the dose in these patients because the tolerability of this ADC will be much better. And I will show you some data later that supports that. And finally, we changed the conjugation, we went from a specific conjugation approach on [ cystine 220 ]to GlycoConnect, which we're using a 601 and 701 as well. And this mitigates the heavy live chain instability that we saw at MEDI3726. So altogether, we believe that we basically changed every component of the ADC in such way that we think we have a very strong candidate for PSMA targeting right now. And the pre-clinic -- sorry, before we go there, this is the ADCT-212 structure. So on the left, you see the antibody with the GlycoConnect remodeling. And on the right, you see the new payload PL1801. So here, we've made the change from the highly potent PBD dimer SG3199 to the lower potency PBD SG2000. And you can also see that we still incorporate here the BCN-Hydraspace structure to mitigate or to offset the off-target toxicity. So compared to the first PBD payloads that we're using, we now made these 2 important changes around the spacer and the PBD dimer itself. This slide shows you that indeed they're still work very nicely. So on the left, you see xenograft model, prostate cancer xenograft model. And you see that a single dose of 5 or 10 mg per kg is still very active. Of course, the dose is slightly higher than what you used to see with our PBD-based ADCs because the PBD dimer itself, as we're using it right now, is slightly less potent. So you have to give a little bit more, but still 5 or 10 mg per kg is a very reasonable dose. Single dose activity, very good tumor regression in this model compared to the nonbinding control ADC in purple. And also, the toxicity profile has changed considerably. So now we can dose up to MTD, which is around 20 milligram per kilogram of this ADC in the red. And you can compare that to 2 milligram per kilogram for the previous candidate in red. So we can dose about tenfold higher of this ADC in rats, and that is accompanied by a very nice PK profile as shown in the graph on the right. So very high nice PK, stability very good, all looking very nice. And this, in the end, means that the therapeutic index for the new candidate has gone up from 4 to 10. So a quite significant improvement preclinically at least in therapeutic index, if you compare it to the previous candidate MEDI3726. And I think this nicely shows that how we are continuing to improve on our PBD chemistry and how did results in even better candidates for the future. And as you may know, we still have quite a few targets to go in our PBD portfolio. And of course, we will do the same going forward with those candidates. So that brings me to the end of my slide. So I'm happy to go back to Chris and...

Thank you very much, Patrick. It's a real privilege to work with 2 such innovative productive and dedicated scientists and clinicians as the two of you. It's a real privilege every day. And thank you very much for the dive you've given us into the ADCT solid tumor pipeline today. It strikes me that they are really innovative drugs. Each one is highly differentiated. If you look at the different mechanism of action of Cami, for example, operating on the regulatory T cells, or you look at KAAG1 as a really unique first-in-class target. But also you're addressing areas of very substantial unmet need. We're looking at lung cancer, ovarian cancer, prostate cancer, head and neck, sarcoma. So areas where patients have few effective patient options. There are many patients with relapse and have very poor prognosis. And they're very large patient populations as well. So these are important opportunities that we're looking at. It's also been absolutely fascinating for you to show us how you optimize the therapeutic index of these molecules. How you can tune each individual component of an ADC and the ADC as a whole to optimize that therapeutic index and increasingly provide these targeted therapies, which are well tolerated by patients and effective in hard-to-treat relapsed/refractory tumors. So this next generation of ADCs is really looking very promising. And behind that, you've continued to create the toolbox of technologies which we can use for our research pipeline in the future, which I very much look forward to. So with that, I look forward to updating you all with these data as they become available over the next 12, 18 months. It's very exciting for us. And I'd now like to open the line for questions and hand over to the operator. Thank you. Operator, can we have the first question, please?

Yes, sir. Our first question or comment comes from the line of Tazeen Ahmad from Bank of America.

And thank you for hosting, that's been super helpful. A few questions from me. Can I -- maybe with Cami. So you guys are enriching your pembro and Cami study. And I just wanted to get a little bit more color from a comment that was made in your prepared remarks about that this combination could fill in the gaps where pembro is ineffective. And I'm just wondering can you give us a little bit more color on what that gap is? And potentially, I know it's difficult to give us a sense of how big that market opportunity could be? And when would be the next time we would see data from this study?

Thank you, Tazeen. Joe, can I ask you to?

Yes, fill in the gaps the term I use because I know that there's patients who relapsed after PD-L1. And the gap is that PD-L1 doesn't have any effect on the PDL. PD-L1 doesn't have any effect on the Treg. So it's that other mechanism that comes into play. I mean the immune system has lots of ways of responding and adapting. And so it's the way that we get around the fact that in these relapsed patients that this new mechanism might be available. I think it's -- and sort of a rule I use -- well, there's a couple. I don't estimate the market size because I'm the medical person. However, one of the things that happens with cancer treatments is the more patients that get a new therapy. Unfortunately, the more you learn about relapse. And so I -- what I -- what we like about this is this combination makes really good biologic sense. We can test it really well with biopsies. We can enrich the population. And we can really add to the response rates from a very good drug, it's widely used. So that all adds up to very positive medical impact.

Okay. Any idea about how big the market opportunity is just that we have a sense? And then when is the data due?

Yes. Chris, I'm going to have to pay us on the market opportunity because I just don't -- that's not medical for me, sorry. But I mean, I look at it as a large population of patients who were on PDL therapy. I mean with that market opportunity, as you know, that's broad-based, that this is going to be a broad-based therapy because it's not tumor-specific. It's mechanism-specific, okay? Having said all that, we're getting close to the end of the dose escalation. We're allowed to do some expansions, but I can't tell you when we're going to have data next because it depends really on when we see responses, how many patients we want to enroll before we feel comfortable. And I don't want to speculate on a date until I know that we can make a date. And we're still in the area where there's a bit of uncertainty about what may happen next in terms of how many patients we need and what we see. So when we know something, we will make sure it is disseminated appropriately.

Our next question or comment comes from the line of Matthew Harrison from Morgan Stanley.

Speaking for Matthew Harrison. So I want to ask about the as well. So I want to ask, what's your view on what type of responses you need to see to move the pembro ahead?

Sorry to move.

The responses to move pembro ahead.

To move pembro ahead?

Yes.

To move Cami pembro ahead. This is unexplored for us. And so I can't really tell you a number. It depends on the patients. It depends on the -- it will depend on the tumors that we choose to expand, depends on the refractory status. There really isn't a number yet. But as you probably know, just to give you an example, the response rates for pembro in ovarian cancer, just pick one of the cancers that we are looking at, are in the range of 8% to 10%. So you can think about number above that for this combination that would be really significant. That's just one example. So just I don't -- we have to adapt as we see data and we have to adapt as we choose tumors before I say here's our decision.

Our next question or comment comes from the line of Brian Chan from Cantor Fitzgerald.

And thanks for putting this event together. I have a question on your AXL-targeted ADCT-601 program. So we get a lot of questions on the amplification selection from investors. It's great to see the visibility that you have in the program today with the Arm A and Sarcoma AXL amplified tumors. Can you give us some guidance on how you're selecting the AXL amplified patients? any color on where you differentiate in terms of the screening assay that you'll be using compared to others? And then I have one more follow-up.

Thank you, Brian. Patrick, maybe you'd like to talk to the assay first and then, Joe, perhaps you.

Well, yes, it's a clinical thing. So I'm not entirely sure which assay they finally decided to pick, but it's one of the standard Foundation Medicine.

Yes. And there's actually a -- yes. And there's a library of patient data that we can choose from. But I can't give you really the specifics of the assay at this point in time.

Because you were referring to the assay to measure the gene amplification, right?

Yes. So -- and also how are you thinking about the cutoff? I think that the threshold to be seen that as AXL amplifies still a little bit hand waving You think -- I don't know if you have any insights on how we should think about what should be classified as AXL amplify tumors?

That is a really good question. I'm pretty -- I'm sure we can do better than hand-waving, but we have to do some learning and validation here. So we have to see what happens with -- you know as well as I do. You have to look at what happens with level of amplification and that little bit of amplification. We can't -- and actually, the kinetics here can actually do that kind of modeling for us once we have some patients. So ultimately, once we have some data, we'll be able to choose a kind of a cutoff. And as you also know, the regulatory agencies often have some input into how you're deciding on which patient may or may not qualify. So at this point in time, we still -- we still have to look at some data before we can give you an answer like to that question.

Okay. And one more question on your manufacturing front. So I recall that you were looking at a new lyophilized formulation for the AXL program. I'm just wondering if you can give us an update on where you are and how that overlap with the -- your expansion cohort?

Correct. I mean we have reformulated AXL or -- sorry ADCT-601, into a lyophilized product. The Phase Ia was done with a liquid. We now reformulated into a lyo that's ready for the start of the Phase Ib. In fact, it's ready. So from now onwards, we will use the lyophilized product in the clinical development, and that can also be the commercial product if we make it to market.

Our next question or comment comes from the line of Boris Peaker from Cowen.

Great. So my first question is on Cami. Can you maybe discuss what is the mechanism of the Guillain-Barre syndrome? And why do you think the mechanism may be active in liquid tumors, but not solid tumors?

Patrick, I think you can do that.

Patrick?

Well, to start with -- of course, we don't know exactly yet what the mode of action is for the existence of Guillain-Barre and Hodgkin. It's clear that we only see it in Hodgkin lymphoma patients and not in any other indications. So we haven't seen it in non-Hodgkin lymphoma in AML or on any of the solid tumor patients that we've been dosing so far with Cami. So the hypothesis we're working on is that, of course, we deplete Tregs with Cami in Hodgkin lymphoma patient as well. And there is a theory that some patients may have preexisting autoantibodies against the myeloid shield. There may be various reasons why they have them. But at least the idea is that if you would deplete Tregs and these patients then become activated and start to produce more of these autoantibodies, again, the myeloid shield that will finally cause the Guillain-Barre. So this is a hypothesis. We're currently trying to bottom it out whether this is indeed the case. But of course, it's likely that it's somehow linked to the depletion of Tregs in the Hodgkin population specifically. But the mechanism behind that, in the end, and what then gives is the Guillain-Barre is still an area of research for us.

Got it. And my second question is on the ADCT-212. There's a lot of competition in targeting and various approaches are being used. Just curious, where do you anticipate to develop this? How do you think this will fit into the treatment paradigm? And how do you think it will differentiate from other approaches?

Joe, you will take that?

Well, I can respond to that. We -- first of all, we're still in the early stages here. So I want to get it into the clinic, get through Phase I, et cetera, et cetera. But I mean, my sort of view of of prostate cancer is in the metastatic realm, I don't think that there's anything that is not going to allow for a new competitive product. And that's what I think this is going to be. So it's -- and as you know, one of the ways cancer treatments progress is that new drugs come -- become available. And then instead of patients going on to a place where there's no new drugs, they get these new drugs. I mean I look at the landscape for prostate cancer, the aging population, the detection and all of that and just say there's definitely there'll be room for a new drug like this. And we already showed you that we improved the antibody, we improved the payload, we've got all these things that are very positive here. I just -- I don't think about too much that there's not going to be room. So that -- but like I said, I like to get into clinic. I'd like to get through Phase I. And then I'd like to find, okay, where exactly could this make an impact? But I just don't think there's going to be a problem finding a part -- a spot for this.

Our next question or comment comes from the line of Tazeen Ahmad from Bank of America.

So I did need to ask you one question about 6 months. So is it primarily the case that your HydraSpace which eliminates a lot of the off-target expression. Is your view of how you mainly differentiate from other AXLs that are in the market or trying to be developed? And I was just wondering what your thoughts were about the [ gema ]molecule that was discontinued and if there were any learnings from that?

Well, of course, I mean, having less off-target toxicity is quite key. But of course, in the end, it's all about efficacy as well. So I think our PBD molecule really differentiates from -- for instance, the[ gema ]molecule that you just mentioned, which was auristatin based. So we really think the truly differentiator in the end is the PBD chemistry and the mode of action of the PBD. And the fact that we're using this linker, which has less off-target talks, of course, is a very important step. But in the end, the ADC has to work in patients. Otherwise, you don't have a drug. So -- but I think that will be driven by our PBD warhead. That's really our differentiating factor in this case. Regarding[ gema ]I really can't comment. I don't know why they stopped the study. It's their study. So not much to comment from my end on that.

Our next question or comment comes from the line of Kennen MacKay from TBC Capital Markets.

This is Kennen MacKay with RBC Capital Markets. I had a candidate specific question and then more of a technology question. On 601, it seems like the synergy with [ gema ]is very exciting. I'm wondering if you expect to see single agent activity there sufficient to advance that as a monotherapy? Or will synergy need to be an inherent part of that mechanism? And then more broadly on the platform, Amgen is giving their business review this morning as well. They just talked about a lot of the low-hanging fruit, so to speak regarding biologic targets having sort of been hit so far. And they give that statement to pitch multi-specific medicines like bi-specific antibodies. But I'd love to get your perspective on how your payload and conjugation platform solves some of the therapeutic index problems associated with some of the remaining targets that are out there as well as whether bi-specific targeting could become a part of your strategy in the future?

Patrick, do you want to start with the gemcitabine combination?

Well, I think it was more a clinical question.

I'm happy to do that one. Just I'll remind you what we're doing in that -- in our dose escalation is a combination in sarcoma because their synergy is really impressive. And the single-agent activity of 601 we saw some stable disease. We saw a partial response, but that's a perfect place where you want to have -- add a synergistic molecule. So it's a setup for sarcoma for us to use gemcitabine combination. So that's the one thing. However, what we couldn't do in our Phase Ia is actual amplification. So in the study that we will start later this year, we're going to be looking at AXL and amplification. And in that case, the single agent may have more activity. Keep in mind, I mean, there's a theme I've been sort of reciting throughout the clinical part of this presentation, which is you want to identify patients whose tumors will respond to your mechanism. And in this case, the mechanism would have to include sufficient AXL expression for the PBD to be able to get into the cells. So we're really testing I think both sides of your question, which is, gemcitabine a good combination? Yes, it is. We already know that it might work in monotherapy? Yes, it might, but it has to be AXL amplified. And this is really a very -- I think, very sophisticated clever protocol, it's a clinical team designed. And I think it covers your question actually pretty well. The bi-specific question was the other one.

Yes, I'm trying to...

Not necessarily bi-specific. But just wondering, again, with a lot of the targets that are easily drug-wise standard monoclonal antibodies or small molecules having been targeted in a lot of the remaining targets have to have very hard therapeutic indexes or in some cases, being undrugable either with a small molecule for monoclonal. Really wondering how the -- your platform sort of addresses some of the remaining targets? so again whether using approaches is something that you've been thinking about?

Ken, I'll start off with that, and then I'll hand over to Patrick. I think it's something that I've been hearing about ever since we started working with antibody drug conjugates. There was, I think, famously, originally only 30 tractable targets. And -- but then if you look at a program like KAAG1, you see there very novel target, very clean in terms of its tumor expression and healthy tissue expression. So you could describe that as a low-hanging fruit in terms of being an ADC target with those kind of characteristics that you're looking for but it was a hidden fruit, which thankfully others working with Patrick have discovered and elucidated. So I'm sure there will be more targets like that. But then you're absolutely right. There are targets where the healthy ticket expression is higher, the expression on some sensitive tissues is high. And I think that's where the toolbox that Patrick has been developing over the past years really comes into its own ability to add another layer of specificity with the tumor microenvironment. For example, the pH of the environment, as Patrick alluded to, in the silent linker system and other approaches in terms of modifying the CDR regions that it only binds in the tumor microenvironment or capping it. So I think there are a whole raft of tools that we have available that we are working on in research, which allow us to add a layer of tumor microenvironment specificity on top of the fundamental ADC delivery approach. But then going beyond that, they're the multimodal cytotoxic approaches. And maybe, Patrick, you'd like to elucidate on those a little bit?

Yes. Although, I mean, it's basically all similar to what Chris has already explained about how we can improve the antibody and make it more specific for the tumor. You can do the same basically. I mean we're building up now our knowledge about how you can make the release of the toxins more tumor-specific or you can make -- or design the toxins in such a way that they won't be active in healthy tissue, again, playing with the pH properties of these payloads. So there are a lot of things emerging in our R&D tool box that in the end will allow us to make them really more tumor-specific and less toxic for healthy tissue. And well, of course, that in the end, they may up a technology or provide a technology or a couple of technologies that will allow you to target those low-hanging fruit targets. That will certainly happen.

Our next question or comment comes from the line of Kelly Shi from Jefferies.

Thank you for offering a very comprehensive presentation. My first question is about 601 program targeting AXL. I'm wondering, besides selecting patients based on AXL expansion level, do you also prioritize the subtypes of sarcoma because there are more than 100 subtypes and probably a variety of AXL expansion level across subtypes?

That's a very good protocol question. We're certainly not going to have 100 subtypes, but we're trying to be somewhat permissive here. Unfortunately, I can't recite the ones that I know are going to be included. But we know based on literature that some or more likely to express AXL. So that's what we're going to do is try to include those ones. I just can't recite it for you because I don't have it in my head. But that's a key to the program for us to try to maximize that expression. So you're 100% right on that.

Great. Also I have a follow-up question on Cami. So the CD25 also expressed cytotoxic T cells beyond the regulatory T cells. And do you concern this might compromise efficacy?

Yes, that's a very good point, and that's something we hear quite often that people are worried about the CD25 expression on T effector cells. But the data, as I showed you and some more data we have and others have shown as well is that there doesn't really seem to be the case. So the CD20 -- well, it doesn't seem to really be the case. Cami or any other CD25-targeting antibody doesn't really seem to impact the T effector cells. So you can tell from the data that I showed you that you really see nice depletion of Tregs, but you don't see that the T effector cells go down. whether preclinically or in the clinic. And we believe it's true to -- it's due to the fact that CD25 expression on T effector cells is there, but it's quite low compared to the high level of CD25 on the Tregs. So you get a sort of preferential binding of Cami to the Tregs versus the T effector cells. So you're absolutely right. It is expressed on T effector cells. But so far, the data doesn't seem to suggest that you also impact T effector cells or deplete T effector cells.

Okay. Super helpful. And lastly, FDA released a guideline for ADC development in this wake. Is there any impact to ADC's early-phase pipeline candidates? And do you have any plans to expand the PK/PD status based on the guideline?

The -- we actually had a chance to look at the guidelines that they released. And our view is that they pretty much codified what they've already been telling us and we've adapted our dose finding in our kinetics and clinical pharmacology to incorporate most of this guidance. So we're prepared to be able to follow it. But as I said, it codifies with what we already knew. So we feel -- we're confident that we can proceed as planned.

Well, thank you very much, everyone. It's been a pleasure to be able to present our solid tumor pipeline and our platform to you today. We very much look forward to updating you as the data becomes available and particular thanks to Joe and to Patrick for your contributions being fantastic, really interesting. So thank you, everyone, and goodbye.

Thank you.

Thank you.