Hey, good afternoon, everyone. I'm Matt Biegler, I'm an analyst here at Oppenheimer. Please introduce our next company, which is I-Mab Biopharma. We've got Sean Fu, CEO, and some of his colleagues to tell you about the story and oncology, which I think is rapidly evolving. So I'm certainly looking forward to hearing about the latest updates. Sean, I'll let you take it away.
Thanks, Matt. So very happy to have the opportunity here to introduce I-Mab. My name is Sean Fu, CEO of the company. During the presentation, I'll be making forward-looking statements, and I refer you to our legal statement. I-Mab is a clinical-stage precision oncology company. We currently have three assets. All are in clinical stage. All are clinically active. I-Mab, as a company, went through significant transformation last year, which positioned us for long-term growth. One of the things we did was implement a divestiture where the company separated out China business and U.S. business. And as a result of the divestiture, I-Mab is now a U.S.-based company with all 24 full-time employees based in the U.S. All asset rights reside in the U.S., EU, or Japan. We have no China commercial interest. We significantly upgraded this executive leadership team.
Today, with me, we have CMO Phillip Dennis and CFO Joe Skelton. This is a team of leaders well-respected and recognized as someone best in their own fields, be it science or business. Obviously, the most exciting thing of our company, the value creating root of our value creation, is going to be the differentiated pipeline. Just give you a high-level rundown. The first molecule is a Claudin 18.2 4-1BB bispecific, and we believe it's a potential best-in-class molecule in the space. It's been developed as a bolt-on therapy to IO and chemo for frontline gastric cancer. It showed clear monotherapy activity in heavily pretreated patients.
I think the more interesting thing is that when combined with IO and chemo, this molecule, we call it Giva, demonstrated a favorable safety profile and markedly improved in objective response rate, ORR, in the dose escalation study relative to the ORR reported in phase III studies that led to the approval of nivo chemo or zolbe chemo in frontline gastric cancer. The second molecule in our pipeline is uliledlimab, or uli. This is a CD73 antibody designed to block the adenosine pathway, which has been developed for first-line metastatic non-small cell lung cancer in combination with IO and chemo. Our business partner, TJ Bio in China, licensed out uli's China rights to Sanofi. We view this deal as a strong validation of the molecule. In the meantime, it also allows us to monitor the ongoing randomized phase II study in China conducted by TJ Bio and Sanofi.
In that study, they are comparing uli versus uli plus IO. So we are positioned to launch our own clinical development program pending a positive clinical readout from the China study. The third asset is a PD-L1 4-1BB bispecific antibody. We are developing it for patients with solid tumors that are resistant to prior line of IO therapies. So we're very excited about the differentiated pipeline. Like I said, all of them are clinically active. To walk you through the highlights, I'll hand over the stage to Phillip, our CMO. Phillip.
Thanks, Sean. It's a pleasure to go over our pipeline, which is very exciting. So let's please move to the slide that starts with Giva. So Giva is our lead program, which is a fairly recent development, and it's based on emerging data from the last two or three months. Giva is a potential best-in-class Claudin 18.2 therapeutic. It's a bispecific targeting Claudin 18.2 as well as 4-1BB. The structure of the molecule is shown in the diagram on the left, which shows there's two binding sites for 18.2 and two scFv binding sites for 4-1BB on T cells. There are a couple of unique features. One is that the Claudin 18.2 is highly avid. If we look at the lead and reference antibody in the class, zolbetuximab, that has an indication of frontline gastric cancer. Our molecule, Giva, has about a tenfold increase in affinity for 18.2.
But what's interesting about the 4-1BB part of the molecule is that its activation of T cells was constructed really to not maximally activate T cells, but rather to act as a partial agonist. This has the intent to lead to localized T cell activation in the tumor microenvironment to minimize known 4-1BB asset class toxicities such as liver toxicity, as well as systemic immune responses. So our line of sight is to develop Giva in frontline gastric cancer in combination with standard of care, as Sean just described. And we are really at the head of the class in terms of testing this new indication by adding it onto standard of care, namely IO plus chemo. Next slide, please. I'm going to start with a review of our phase I monotherapy data with Giva. This is in, again, heavily pretreated patients.
The data that I'm focused on here are 43 patients that was part after the dose escalation. We had a dose expansion looking at gastric cancer patients. This data was most recently presented at ESMO just a few months ago in 2024. This included 43 efficacy evaluable patients that were Claudin 18.2 positive. Of note, in this study, in our monotherapy phase I, as in every study moving forward, our threshold for Claudin 18.2 is about as low as it can go, 1% of cells at 1+ or greater intensity. This is to provide the maximum number of patients who might benefit from Giva. In this phase I study, we observed seven confirmed partial responses with an ORR of 16%. Importantly, in the Claudin 18.2 responders, they ranged from 11%-100%. We're seeing responses in the very low end of Claudin 18.2 expression.
Of the seven responders, five of them had received prior checkpoint inhibitors. Giva not only had notable clinical activity in heavily pretreated gastric cancer patients, but in fact, it was very well tolerated, which is shown on the next slide, please. Importantly, the bullets on the right show and state accurately that there was no dose-limiting toxicity encountered. We did not reach a maximum tolerated dose. We based our dose escalation and ceasing of the dose escalation based on pharmacodynamic markers. And once we plateaued in the pharmacodynamic markers, we saw no reason to escalate further. I'll draw your attention to the chart on the left, which really delineates the treatment-related adverse events occurring in more than 5% of the patients in the study. If you look at the columns for grade four and five, there's only one grade four TRAE, which is thrombocytopenia, platelet count decrease.
Of note, this was not associated with bleeding, so it's really an isolated lab value of limited clinical importance. Importantly, if you look at grade three treatment-related adverse events, that column is notable for the fact that all of the adverse events listed, they're all less than 10%. And if you look at the AEs from 7% to 10%, they're all lab values. They're not clinically meaningful. The clinically meaningful AEs, such as nausea, vomiting, decreased appetite, they're all at a remarkably low rate of 2%. So Giva monotherapy was very well tolerated. Next slide, please. Now, if we compare our data in this phase I study against the early studies from zolbetuximab, again, the first 18.2 asset to gain an indication in frontline gastric cancer, you can see that our threshold for Claudin 18.2 in the top row was as low as you can go.
In a similar population, zolbe had a 0% response rate. Even when they tightened their threshold to 50% of tumor cells staining at two plus or greater intensity, their ORR was only 9%. This shows that Giva compares very favorably to zolbetuximab. Especially if you look at important AEs such as nausea and vomiting, where our rate was 2%, their rate was approximately tenfold above that in the 20%-25% range. In a head-to-head comparison, although it's not been done in a single clinical trial, our data compare very favorably with zolbetuximab. Next slide, please. We've also looked at other assets in the class, and there's a lot of attention now in antibody-drug conjugates, such as CMG901, that targets 18.2 as well.
What’s characteristic of the ADC class, and I would include in that the T cell engager class through molecules such as CD3, that they come with greater efficacy. The ORR for CMG901 listed in this slide from the data presentation was a 33% ORR rate. But importantly, this increased efficacy comes at a cost of toxicity. Here’s showing grade three or greater neutropenia as well as vomiting. So it’s a package where greater efficacy comes with greater toxicity. So when we looked at this data, we basically hypothesized that of any real asset in the class, that Giva is positioned exquisitely well to be combined with frontline standard of care, that is IO plus chemotherapy, more specifically nivolumab plus FOLFOX. So next slide, please. The next slide shows the study that really led to the elevation of Giva as our lead asset.
This is a dose escalation study, again, combining Giva with nivo plus FOLFOX, three different dose levels. We are not modifying the standard of care at all in terms of dose reduction. We're not changing the administration schedule. These patients are all frontline, unresectable, HER2 negative, and have a very low basal level of Claudin 18.2 expression. The dose escalation portion of the study shown in blue is complete. We had 17 patients on this portion of the study. In this part of the study, we did not encounter a dose-limiting toxicity. And again, based on pharmacodynamics, a maximum tolerated dose was not reached. More importantly, when we looked at the data, we found that the toxicity profile looked very much like nivolumab plus FOLFOX.
When we looked at the objective response rate, we saw a marked improvement over the ORR reported in the key phase III studies that led to the approval of nivo plus FOLFOX, as well as zolbe plus FOLFOX. This marked improvement in objective response rate was also accompanied by responders that would not be expected to respond to the standard of care, such as nivo, FOLFOX, or zolbe FOLFOX. What do I mean by that? What I mean specifically are patients that are PD-L1 negative or Claudin 18.2 at 1%, very low levels of each of those biomarkers. If we took the ORR and these notable responders together, we took a look at our planned dose expansion in this study, and we quickly amended the study to enlarge the dose expansion to 20 patients in each of two dose cohorts.
This will allow us to do two important things. One is to validate the ORR that we saw in the escalation and confirm those responders that we saw that had very low levels of PD-L1 or Claudin 18.2 expression. But what this approach will also do, it should satisfy the criteria from the FDA for dose optimization, such that if we replicate the data that we've seen in the dose escalation in expansion, we would want to move expeditiously to a phase III study, a randomized phase III study. I will mention that the data for the dose escalation part will be presented at a major meeting later this year in all of its full data set. The dose expansion data, I will mention that we are ahead of schedule in recruitment.
We have several investigators, known GI oncologists in the United States, who've recruited quite well into the dose escalation. They've seen what Giva can do. They've seen their patients benefit. They're very interested in dose expansion, and through their efforts and others, we're recruiting ahead of schedule for the expansion. We expect the expansion data to be presented in the first half of 2026. Next slide, please. This slide shows the Claudin 18.2 gastric cancer market opportunity on left in terms of the cases. Of the 250,000 cases diagnosed each year, about 195,000 are HER2 negative. About 137,000 are Claudin 18.2. So many, many patients need better therapies, could potentially benefit from Giva, and the right side of this slide, it shows the potential opportunity for Giva, and what I'll start with is the only 18.2 asset that has an indication of frontline. That's zolbetuximab.
Zolbetuximab's label is limited in two ways. First, it's limited in terms of Claudin 18.2 expression. For example, its label is limited to those patients who have greater than or equal to 75% of cells staining at two plus or greater. So very intense staining in almost in the far majority of cells. But importantly, their indication is only in chemo alone. For 90% of patients, the standard of care is IO plus chemo. So if we look at the size of the rectangle in the right lower corner, we can see that the current indication of chemotherapy alone used in PD-L1 patients, less than 1%, and Claudin 18.2, greater than 75%, that is about 8% of all frontline gastric cancer patients that are Claudin positive. So that leaves 92% of the space that could be occupied by Giva.
If we look at the area in purple, less than 75%, Giva has a potential to be first in class there because zolbetuximab has limited activity in patients who have lower levels of Claudin expression, and as I've mentioned, we've seen activity in far lower levels of expression. Moreover, even if zolbe moves into the blue area above the current indication, in other words, let's say they do a study to combine with IO and chemo, what I've shown you from the monotherapy data is that Giva has better monotherapy efficacy and a better safety profile, so even if zolbetuximab moves into the area above its current indication, in other words, expands into IO plus chemo, there's a potential that Giva can still be best in class in that opportunity. This is not a small opportunity.
If we look at the revenue generated by zolbe in Japan in 2024 in its first year of launch, the revised figures show that they made $34 million. And Astellas has projected their peak year sales anywhere between $650 million and $1.3 billion. So if we look at the rest of the space around zolbetuximab, we can see that the opportunity for Giva is quite large. Next slide, please. Well, this proof of concept that we're establishing in gastric cancer is only the beginning. So I've just mentioned to you the opportunity in gastric cancer. That's only metastatic gastric cancer. We haven't talked about resectable disease. But I want to talk about other tumor types because Claudin 18.2 is a very prevalent biomarker in GI malignancies, such as pancreatic cancer and cholangiocarcinoma or bile duct cancer shown on the far right.
Each of these has a sizable market potential of $6 billion and $3 billion, respectively. But importantly, the majority of them clearly express Claudin 18.2. And what this offers to us, having established the proof of concept in gastric, is that if we can combine safely with IO and chemo and provide an important increment in clinical outcomes and efficacy, then we could simply look at frontline pancreatic cancer, for example, and test Giva in combination with a standard of care chemo doublet or in cholangiocarcinoma tested in combination with IO plus chemo. Next slide, please. So in summary, Giva, we feel, is a potential best-in-class Claudin 18.2 therapeutic. It is really at the head of the class in terms of being tested as a bolt-on standard of care to IO plus chemo in frontline gastric cancer.
It's got some really interesting properties in terms of its 18.2 affinity, conditional 4-1BB activation that really threads the needle in terms of activating T cells to the point of getting an enhanced response, but not to the point of seeing liver toxicity and systemic immune responses. It's designed for lasting immune responses across a wide range of Claudin expression. Again, our threshold in every clinical study we do is one plus of cells at one plus or greater. And the response rate and tolerability right now stand out versus the benchmarks. And then finally, I'll just reemphasize that we will present the escalation data in the second half of this year and the expansion data in the first half of next year. Next slide, please. I want to spend a couple of minutes talking about our other assets.
The first asset is Uliledlimab, which is a potential best-in-class CD73 therapeutic. Simply put, it is a very good CD73 antibody. When we look across the bow at the leading antibody in the class, oleclumab, that's now in phase II and III studies with AstraZeneca, there's clearly preclinical evidence showing that uli is a better inhibitor of CD73. Next slide, please. I won't get into this detail, but there are other targets in the CD73 pathway, the adenosine pathway. The key thing that CD73 catalyzes is a conversion from AMP into adenosine. That adenosine provides an immunosuppressive fog around a tumor that prevents an immune system from recognizing the tumor as being foreign and eliminating it. Next slide, please. One of the advantages of uli is how it binds to CD73. Very simply, it binds in a different manner than oleclumab.
What's shown on the next slide? Why does this matter? Because uli is a very straightforward inhibitor. The panel on the left, you can see that as you dose escalate uli, you get complete inhibition of CD73, as well as the inhibition is sustained across concentrations. In contrast to oleclumab, you get incomplete inhibition of CD73, only about 90%. As you dose escalate, you see a paradoxical loss of inhibition of CD73. That is a so-called hook effect. Does this exist other than in cells in vitro? Next slide, please. This is a mouse experiment showing on the left, uli in a dose-dependent manner in a mouse model shows that with increasing dose, there's greater CD73 inhibition and greater tumor inhibition shown on the lower left. In contrast, oleclumab does not lead to dose-dependent CD73 inhibition, and nor does it lead to dose-dependent tumor inhibition.
So we don't know if the hook effect that we see consistently with Oleclumab in preclinical studies will play out clinically. But we know that the data to date with Oleclumab is very modest. And we remain optimistic that uli, by being a better CD73 inhibitor and by virtue of the fact that our strategy is to select patients based on CD73, again, uli has great potential to be a best-in-class CD73 asset. Next slide, please. So this is the data that really got us excited, where this is a frontline study in non-small cell lung cancer, where patients were treated with uli plus a checkpoint inhibitor, toripalimab. Patients were divided into PD-L1 status, all comers or greater than 1%, and CD73 high versus low expressors. And this is what's distinctive and differentiated about our approach.
We are developing this novel IO agent by trying to identify those patients by measuring the target of uli in the tumor. So if we look at the CD73 high population, what we can see is that the ORR is more than double that of the CD73 low population. And if we look in the PD-L1 greater than 1%, yes, it's threefold greater. But importantly, it's also greater than a reference study, KEYNOTE-0 42, which is pembrolizumab monotherapy in the same population. So we're very excited about this. And as Sean mentioned, this data is being extended and replicated in a prospective randomized study of uli plus IO versus IO alone in a CD73 positive population.
And positive data from that study will be a trigger for us to continue the development of uli subsequently. Next slide, please. I'm just going to finish with the responses to uli, which are long-lived.
This is all data that's available. Next slide, please. Let's just go to ragi. I'll just give one slide summary, which is a very interesting PD-L1, 4-1BB bispecific. It's got a similar structure shown on the left with two binding sites for PD-L1, the same interaction with 4-1BB. Again, it's meant to lead to conditional activation. And if we go to the comparison with the acasunlimab, please, this is the results of our phase I study. We had an ORR of 27% in highly pretreated patients, including one patient that had ovarian cancer who was receiving this as seventh line of therapy. She had a complete response to ragi. In contrast, if we look at the lead PD-L1, 4-1BB bispecific, the ORR was 7%. And even at higher doses, it was only 13%. So we compare favorably in terms of ORR against acasunlimab.
But we are cognizant of the fact that our rates of liver toxicity here at 25% was greater than that observed with acasunlimab. So what we've done is basically recognize that the liver toxicity is potentially related to too much stimulation of T cells. So we've now amended the phase I study to add cohorts where we're decreasing the dose of ragi and separating out the time between cycles to maximize the therapeutic index. So next slide, please. I'll just then summarize and say that we have important readouts coming up here. As maybe inquired about later, we clearly have cash available to extend our development into 2027 and reach these key clinical milestones I've described. We have important readouts in the second half of this year and first half of next year for Giva.
That randomized data coming from China of uli plus the checkpoint inhibitor, we expect that to be available in 2026 again, which could be a catalyst for further development. And then we have an ongoing phase I study with ragi. And we expect these additional cohorts to be available later this year. So thank you. I'll stop there.
All right. We've got maybe four minutes left. Super insightful presentation. I think last time we spoke maybe four months ago or so, it seemed like uli was the lead asset. And now you've kind of swapped them with Giva. Was that maybe driven by the data that you saw at ESMO or maybe other things that have happened in the competitive space since that you think that this is kind of your lead asset now?
No, I'll let Sean weigh in as well. But this is shaped by our data.
Our Giva data that's unfolded, really, is making us stand up and take notice. We have not lost any confidence in uli. We have great confidence in uli. It's just the fact that the data coming out of this phase I in combination with nivo and FOLFOX is so strong that we want to put everything we can into it and simply be very rational and logical and let the development of uli follow with a positive signal that comes from our collaborators doing that randomized study. So it has nothing to do with uli's evaluation. It has nothing to do with oleclumab data. We're a better CD73 inhibitor than oleclumab. I have great confidence in that.
But we're simply focusing our development right now with the asset that has the highest probability of success where we're generating data in real time that is simply building on the positive signals that we're seeing.
Makes sense. And I guess so the value prop for Giva is that vis-à-vis ADC, the efficacy might not be there, but the safety is better. So therefore, it's really the only thing that you can compare with kind of standard of care, which is nivo plus chemotherapy, right?
I think so. I think to combine having led an ADC program in the past, a very large one, that to add ADCs with a chemotherapy warhead to a standard of care regimen that includes a chemotherapy complex chemotherapy with IO, something's got to be taken away from the standard of care. We don't have to do that.
We have a good therapeutic index where we have efficacy as monotherapy, but a really tolerable safety profile. So we're just exquisitely suited to simply bolt on to the regular standard of care. So every patient gets the standard of care. And in a randomized study, health authorities are going to love that because every patient gets the standard of care. It's not adulterated. And when you move ADCs into the frontline, often you have to give up something on the standard of care because it's just hard to tolerate.
Yeah, makes sense. Do you think that there might be an accelerated approval strategy here, or are you going to have to run a randomized trial against nivo chemo combination?
I think they're not mutually exclusive. I think, obviously, we will be generating randomized data. That's the next important step with that. However, I think there are paths, especially with Project Front Runner. There's other new opportunities for accelerated approvals that we will be exploring.
Makes sense. All right, I-Mab. Thank you so much for your time, and thanks for the update. We'll speak to you soon.
Excellent. Thank you very much.
Thank you.