All right, I'm Stephen Willey, one of the senior biotech analysts here at Stifel. I'm glad to have with us here for the next session, CEO of Zymeworks Ken Galbraith. Ken, this is just going to be kind of a casual fireside chat. I know you excel in that format. I'm not sure if there's any opening statements you want to make before we get into Q&A at all?
Not really. Just to say, we worked really hard since the Jazz deal completed at the end of 2022 to put ourselves in the company exactly in this position that we're in now with Zyme at a regulatory review stage, and we're right in the front of putting five of our medicines out of the five by five into clinical studies in the next 24 months. So we're really excited about where we've positioned the company, and we just need to execute over the next 24 months to really get to the company's potential.
Okay, great. So I think we're going to start in with some Zyme questions, and then I know you just had a busy AACR, so if I guess we have some time, we can touch upon some of that. So I know HERIZON-GEA data before the end of this year per your partner Jazz. This is probably the catalyst that we get the most amount of inbound questions on currently. So KEYNOTE-811, both kind of with and without Pembro is kind of the competitive benchmark everyone's looking at. There's certainly some nuanced differences between your trial and KEYNOTE-811. So I guess maybe you can just start us off by speaking to some of those differences in terms of trial design and how you think those differences could potentially influence outcomes, and then also maybe labeling and utilization.
Yeah, absolutely. I think it would be hard to make some direct clinical trial comparisons between KEYNOTE-811 and HERIZON-GEA-01. Obviously, Zyme is not Traz. As we've seen from our clinical data, it's a completely different mechanism with the bispecific antibody that we have versus Traz . So that has so far in the clinic turned into some different clinical responses than you might have expected with Traz . Design is quite different. In the KEYNOTE-811, it's a triplet for every patient. In HERIZON-GEA-01, we're studying both Zyme plus chemo and then Zyme plus chemo plus PD-1, which I think hopefully gives us a better data set of when PD-1 might be applicable in this patient population because I don't think it's for every patient. But in KEYNOTE-811, it had to be for every patient.
Difference between Pembro and Tisla on the PD-1, I'm not sure I'm smart enough to know what the differences are between those two and whether they've really been studied head to head. Obviously, a much different population; ours is broader. We chose to put the esophageal adenocarcinoma patients in with the gastric and GEJ, and I think there's a good reason to do that given the unmet need for that population. We have a pretty global study. I think one of the advantages that came to us of not having U.S. sites is making us work harder at having a more global study and a much more diverse population as a result. There's over 300 sites still enrolling right now in that study.
But I think the patient population comes out of that might be more diverse because of the need not to be in the United States because of the accelerated approval with KEYNOTE-811 a couple of years ago, and maybe that decision we could have had the sites, but I think that'll be different for sure. I think there's differences in the way we think about PD-1 status. We're not stratifying our patient population for that. I think because we have the three arms, we don't need to on an exploratory basis, we will be able to analyze subpopulation data and disclose it as to the benefit for adding a PD-1 to the Zyme chemo combo given what your PD-1 score is going to be and whether it's positive or above five or above 10.
So I'm hoping that we have a more global population that's diverse, and I think the data set that comes out of that is more complete and diverse, and that'll be meaningful for regulators and physicians to really understand the benefit of Zyme plus chemo and opportunities where adding a PD-1 in a tolerable way may be possible and what benefit can be seen with that.
Okay. I definitely want to talk about PD-1 status as it pertains to this trial. But you mentioned esophageal and the inclusion of these patients. They were excluded from KEYNOTE. When you look at the kind of histology-specific outcomes in KEYNOTE, right, there appears to be a little bit of slippage as it pertains to GEJ relative to gastric, which I think some KOLs have speculated is due to a slightly more heterogeneous level of HER2 expression. And I guess as you kind of walk your way up the esophageal tract, would you expect that to be the case for these esophageal patients? And if you have this representation in Horizon, that might be a little bit more harder to treat. Do you think that kind of ratchets up the competitive bar a little bit?
Yeah, I think what our clinical experience for Zyme has shown us so far is that because of the mechanism that we have, Zyme tends to shine where there is heterogeneity. So it might be because the mechanism for Zyme is so different than what you saw with Traz or Pert or Traz and Pert together that you'll see the ability to even out these differences that you might see in GEJ or gastric or esophageal, which just might be a reflection of those things being treated by something that's not Zyme. So I think if you look at our phase II data, we had a number of responses per patient. I think we had six patient responses in the esophageal cohort, esophageal subpopulation at phase II of the 201 study, which was published at ASCO GI in January 2023, but one complete response.
So from what we saw from our phase II data, and there's a reason why we think we see that between gastric and GEJ and esophageal, we broadly saw a very similar response in phase II in that patient population in the study with chemo. So that gave us some confidence. There's not been a pivotal HER2 positive RCT in the esophageal patient population. HERIZON's the first one. So maybe there's always some risk of being first in a randomized study to include that population. But based on what we saw in phase II, there's no reason to exclude them. And if we can get those patients a HER2 targeted therapy where there isn't really one today that they can benefit from, that's a huge unmet need and much higher in certain geographies than others.
What we're trying to do is give this physician group the ability to potentially treat their gastric, GEJ, and esophageal cancer patients with the same regimen if our data can support that. We feel confident that it will based on everything we know about Zyme and specifically our GEA study in phase II that was completed and presented already.
Yeah, no, that's a really good point about Zyme and the heterogeneity. And I guess there's also some reward to being first as well as risk. So maybe we could talk about PD-L1. So we know that there was this kind of pronounced detriment in the 15% of PD-L1 negative patients within KEYNOTE. I don't think there's been a great scientific rationale offered up for this in terms of detriment anyway. This 85% of PD-L1 positivity in KEYNOTE seems to align with some translational data that Daiichi has shown within HER2 suggesting that the concentric circles of overlap between the two targets is pretty high. Does that 85% align with what you're expecting? And is there a risk to the triplet arm if this proportion of PD-L1 negativity is somehow much meaningfully higher than what we've seen in 811?
No, that's not what we expect to have in our patient population. Again, I can't explain the split in the KEYNOTE-811 study because it's not my study. I don't know all the details of how they ended up there and whether that just happened or if it was enriched to be more on the PD-L1 side. I don't know the answer to that. I think there's two things we know. There's lots of published data about what percentage of patients might be both HER2 positive and PD-L1 positive. The most recent one that was published in 2023 by Meta et al., which was U.S.-based, said it was 4%. That's a long way from the 85% we see in KEYNOTE-811. There's a range of published studies that try to answer that question with different modeling, different geographies.
All we'll really know is when our studies finish, we'll know what the demographics are and the population statistics around that. So our best estimate is it's about 55% will be PD-L1 positive and 45% PD-L1 negative. Of course, we're over a different population too with the esophageal there. There's some differences between gastric and GEJ. But I think where we're recruiting across the globe, which we think is pretty much real world everywhere, despite the lack of U.S. sites is there. We're not stratifying for PD-L1 status. We think Zyme plus chemo, given our phase II data, works regardless of PD-L1 status very well. Will adding a PD-L1 maybe work a little bit better for some patients? That's the reason we have an arm C to study that. So I can't really speculate on the differences. I think we know from Zyme clearly how it works.
We had a three-cohort design, which I think gives us a more diverse and better data site. There's no expectation that Zyme will work differently either with chemo or with PD-L1 in different PD-L1 status, but the data will tell us that. I just can't explain otherwise why they ended up where they ended up. I don't think there's a detriment to PD-L1 negative patients. There's no published data that shows that patients with HER2 and GEA that are PD-L1 negative should do worse than PD-L1 positive. I think if you enrich the population end up with just a 15% cohort, which is 50 patients an arm, and if you stratify it that way, which we're not doing, those small Ns can make the statistics meaningless.
It was detrimental to the study, I think, because they had to take that arm off in order to be able to file on PFS because the OS on that stratified population was negative. But I don't really think that's real. I think that's low-arm statistics, and it's why you need to try not to skew these patient populations when you're screening them. Just take them as they come real world, and hopefully the diversity of your clinical sites and where you're recruiting gets you as close to the real world population as possible and not what's probably an anomaly from our perspective.
Okay. So the competitive inference there is that being able to procure a label in PD-L1 negative patients is a distinct commercial advantage for you, just given that you believe that the size of that market is much larger than what is represented in KEYNOTE. Is that a fair statement?
Yeah. I mean, if our data supports it, we'd like to support a regimen of Zyme plus chemo regardless of population characteristics, including PD-L1 status, regardless of gastric, GEJ, or esophageal. And if PD-L1 provides a benefit for a certain segment of the population, then we'll have the data to support that as well as the tolerability data to support adding a PD-L1 to Zyme plus chemo. So the data will be available to guide regulators and physicians as to where the appropriate positioning of Zyme should be. Right now, we think it's been pretty broadly applicable to all the patient populations we study regardless of all those subpopulation factors. If you look at KEYNOTE-811 data, there are some population characteristics where it doesn't work as well, even in PD-L1 positive patients. Older males, Asian ethnicity, those didn't do as well as the rest of the population.
We would hope to have some good consistency of response across all of those attributes that you'll look for in the population analysis that we'll do and actually disclose.
Okay. And so I know that you've emphasized that within this trial design, either arm can statistically win, right? So Zyme-based doublet or triplet either could secure a regulatory approval. But can you kind of help us understand how regulators will be thinking about the contribution of tislelizumab here, just given that there's not a pre-specified comparison of the two experimental arms?
Yeah, there's no pre-specified comparison between arm B and arm C, the doublet and the triplet, but it's a randomized study. And it's a large randomized study and even larger for the OS endpoint now. So I think, and again, from a regulatory standpoint, FDA is going to be reviewing two different BLA applications. And hopefully, we have biliary tract cancer approved already, so it's a supplemental. And I think BeiGene has expectations of having Tisla on label for another indication. So you're going to have two supplemental BLAs under review by FDA for extended labeling using a common data set, but not necessarily common labeling discussions. So even though there's not a direct comparison statistically between arm C and arm B, that's not the way the hypothesis was set up.
It should be obvious in the data the contribution of adding PD-L1 to a Zyme plus chemo doublet and for which patients you might see a more substantial benefit over Zyme plus chemo. And so until we see the data and until we give regulators a chance to review that data, both in the U.S. and outside, it's premature to talk about it. But don't forget, it is two different regulatory dossiers that will have to be approved and have a label extension and the data set's common. So that contribution of components should be sitting right there to be evaluated if regulators choose to.
Okay. And in the context of the doublet procuring regulatory approval and for whatever reason the triplet not, would you look to generate additional data in combination with Pembro, again, just kind of given that it's somewhat institutionalized at this point o r do you think that physicians would be willing to just add Pembro in the absence of prospectively generated data?
Yeah, I mean, it's a hard question. I mean, it's obviously an approved agent, and it's been used pretty widely by physicians. I mean, obviously, nothing would prevent the ability to study other PD-L1 inhibitors post-approval beyond Tisla, especially if physicians express some desire to use their favorite PD-L1 because they think one is better than the other given their experience. And that does happen depending on the indication. So it wouldn't stop us or an investigator from studying whether there's any differences in response between using one PD-L1 inhibitor over the other. But I presume that's going to be something in a post-commercial world that would be studied.
Okay. I know Jazz mentioned having some interest in locally advanced pre-metastatic disease. This was at their Zyme-focused R&D day event. I know that there's been some small trials of HER2-directed therapy in this setting before, but I 'm kind of surprised that there really hasn't been a registrational effort to leverage HER2 status in this setting. So what do you think of this opportunity and I guess kind of the underlying risk associated with pursuing a label given that a standard of care arm wouldn't include a HER2 targeting agent?
Yeah, I mean, obviously, HERIZON-GEA-01 is just the first study we have ongoing in this patient population. Nothing prevents us from studying other patient populations beyond where we are in the metastatic first line. Obviously, we can look at our data with Zyme in metastatic breast cancer and see that there's the potential for some benefit in an adjuvant or neoadjuvant setting. And so when you would look at this tumor type, you would ask yourself whether that has the potential to also be something that could broaden out the potential patient population for Zyme. And there's pretty good data body of evidence right now. It's just that's something that would be interesting to study post-approval for HERIZON-GEA-01, how it would be studied with the issue you talked about. I think we would want to study it.
I think physicians would like to understand if that could be another potential benefit from Zyme beyond the initial label that might be secured with Horizon GEA-01 study. So I think there could be a whole strategy around looking to broaden out the patient populations to study beyond the initial population supported by Horizon GEA-01 study. And I'm sure that's in Jazz's mind right now about not just what they mentioned, but probably other things beyond that.
Okay. I don't think the pan-tumor label that Daiichi just received really caught anyone off guard, but just kind of curious as to how you're thinking about the competitive dynamic in second-line biliary. And whether or not you would expect a Zyme approval to be limited to those IHC 3+ patients to which the Enhertu label is specific to.
Yeah, I think first, it's a really interesting regulatory strategy that we took note of. And I don't think this is the last ADC that will try to look at this pathway. And for us, at least with folate receptor alpha ADC coming along, which is a pretty broad set of indications, and our mesothelin 2+1 T-cell engager coming along, which is also a broad set of indications. It's interesting just to think about the regulatory strategy that they can utilize to emulate what they've done in the HER2 space. It's pretty interesting the way they've worked with regulators to broaden out the label for patients where there might be an indication of interest. It's obviously accelerated approval needs to be proven. I think our study with Zyme in this patient population is quite different. I mean, it's a much larger data set.
It's a data set which I think has shown very comparable ORR and a better tolerability profile given the patient discontinuation rates here. Obviously, this is more limited for T-DXd because it's IHC3+ only because the IHC2+ response rate was zero. So obviously, there's not the breadth of potential indication, which we saw a lower response rate in 2+, but we still had responses in 2+. So that tells us something. And obviously, this bigger data set we've continued to follow since we put out the phase IIb data at the end of 2022. So I think we have a much more extensive data set, which hopefully we'll get a chance to share this year at a conference or during the course of the year.
And I think there's every expectation that Zyme provides you a high level of durable response and a long duration of treatment and duration response. And hopefully, that provides a meaningful result for patients, which the current data set for T-DXd and the pan-tumor indication just doesn't have the data to support anything like that right now. And they would have to create that data set around it. While they're creating the data set, we're moving from second-line monotherapy to our confirmatory study of first-line patients with PD-L1 and Gem Cis. And so we have the chance to combine in an earlier patient population, which could be pretty interesting based on where we are in second line. And I think that goes beyond where T-DXd has the ability to combine tolerably right now. Maybe they'll find a way with that agent to do it, but we haven't seen that yet.
I think for biliary tract cancer, we feel pretty confident where we are right now with our current data set, which is continuing to mature and get more compelling than even the original data we put out. We have the chance to broaden that in combination with other agents in an earlier line of patient care, which I think could be pretty compelling for us if we can find data to support that use.
Okay. Just a quick question on the confirmatory frontline trial that Jazz has outlined. It looks like the primary endpoint here is PFS and the IHC 3+ population specifically. I guess I know this isn't a big trial. I think it's less than 300 patients. Is there some threshold representation of 3+ patients that you're going to try to proactively enroll into this trial?
Yeah, I don't think Jazz has or we have talked much more about this study other than what's on clinical trials. The study is properly sized for the endpoint that we have. And I think the fact it's not as large of a study as you might have seen in other confirmatory studies in biliary tract cancer with other targeted agents is just because of the expectations of what Zyme in combination with PD-L1 and Gem Cis will bring to this patient population versus PD-L1 Gem Cis, which we know from the TOPAZ-1 data and data supporting the competitor molecule. In the subpopulation analysis, we're not overwhelming for the HER2 amplified patient population. In fact, in TOPAZ-1, it was a negative overall survival advantage, the only subpopulation that was.
So I think the size of the study is structured around the power that we think Zyme will have in combination in this patient population with standard of care. And again, it allows it to be executed in a much shorter time frame than we've seen with other confirmatory studies in biliary tract cancer. So pretty confident in the approach that Jazz is taking here. And I don't think we'll provide more details about this until the appropriate time.
Okay. Maybe just in the last couple of minutes here, we can kind of just touch upon AACR. I know it was a very busy, visible meeting for you guys. I think we've seen a lot of companies or there's a lot of ADC presentations there. I think it's probably an understatement. And I think we've seen companies try to limit risk on the ADC front by either pursuing novel targets with established linker payloads or vice versa. You're obviously now kind of doing the latter with 191 and kind of the rest of the IND candidates that you flagged. But do you think that the clinical validation of 191 and kind of proof of concept established with the linker payload eventually kind of pushes you into more of this novel target space?
Do you have any discovery efforts ongoing at the company on the biological front to kind of unlock newer targets?
Yeah, I think in the 5x5 on both ADCs and T-cell engagers, we have some new elements in our platform. We got a proprietary payload in our ADCs o n the T-cell engagers, a new CD3 that's never been used before. And in our tri-TCEs, some element of building in the CD28 costimulatory factor into the same biologic structure, which no one's done successfully yet. So we felt for the first five, we would work with targets that were more validated and understood. And in some ways, like folate receptor alpha, they're validated positively because Mirv's approved. For NaPi, they're validated by unsuccessful clinical attempts, but we still understand the biology. And for GPC3, you can see there's validation by other product formats, just not by an ADC approach yet.
Mesothelin would be in the category of lots of unsuccessful ADC attempts, but maybe a 2+1 format in a T-cell engager like we have with a novel CD3 concept, it might be successful. Beyond the 5x5, where we have a great developing portfolio now to create two new INDs per year. A fter the 5x5, there's a lot more novelty and diversity around both elements of the ADCs and T-cell engagers f rom a brand new payload, we think is the next step beyond Camptothecin analogs. Dual payloads are a bispecific strategy we presented at AACR. And on the T-cell engagers, different tri-specific formats we think are pretty interesting beyond the current one. And also the ability for us to branch out from our current therapeutic concentration in gynecological cancers, lung cancer, and GI cancers to hem-onc, where we have good experience and would like to look at some areas.
Also even outside of cancer, eventually looking at autoimmune and inflammation, where we have some preclinical assets we're developing as well towards the clinic. So I think as we move from the 5x5 to the next portfolio, which we call Advance, you'll see a lot more novelty, breadth, and diversity there because we will have taken some of the platform risk out of our next-generation ADCs and next-generation tri-specifics. We have an R&D day in Q4. And I think that's where we'll talk a lot more about what's beyond 5x5. Right now, we're focused on helping Jazz and BeiGene with Zyme and getting the first two of the 5x5 in the clinic and naming the fifth one. And then once we've done that, we'll turn our attention to talking more about what you'll see behind that. And I think it's very interesting and compelling.
We're kind of excited to go to that next level. We talked a little bit about it at AACR, kind of a preview, but we'll talk a lot more about it in Q4 once we get to the end of the year.
Okay. Very interesting. We are out of time. But Ken, thanks for joining us. There's obviously a lot of strategic interest and enthusiasm for ADCs and multi-specifics right now. And you guys are certainly kind of at the forefront of some of those efforts, so.
Yeah. Yeah. Then we can confirm there's a lot of interest in folks wanting to partner and collaborate with people who are at the leading edge of these areas as we are. You just got to figure out how that interest can help our shareholders, how we can accelerate things, how we could broaden out what we're doing now. Because with the capital and talent right now, we can execute these programs into early clinical development on our own without the help of a partner, which is a good position to be in. But we'll continue to evaluate interest that we might be receiving and just determine how it's going to help us in building the portfolio that we want to build, not just in the 5x5, but even in Advance, the next part of our portfolio development.
Very good. Sounds like a first-world problem.
Yeah.
Ken, thanks for joining us. Appreciate it.
Thanks very much, Steve. Appreciate it.