Great. I think we can get started. Good afternoon, everyone. Thank you for attending Jefferies Healthcare Conference. My name is Yifan Xu, from Jefferies Biotech Research Team. Today, we are very pleased to have Dr. Sean Fu, CEO of I-MAB, and Dr. Phillip Dennis, CMO of I-MAB, for this session. Welcome. Thank you. To get started, for those who may be less familiar with your story, could you please provide a high-level overview of your company and your differentiated pipeline? Absolutely. Thanks, Yifan. Can you guys hear okay? Mic calling? Okay. It's a great opportunity here to share the I-MAB story. I would say that when you think about I-MAB, it's a company with a long history, but really, it's a new company reborn through recent strategic changes. We had a divestiture last year, separating out assets from China from the US. The U.S. company, which is the LISCO on Nasdaq, now is really focused on clinical development, really focused being a U.S.-based biotech company, and really focused on efficiency. We have 25 employees, all in the US. We are a U.S.-based company. Out of our pipeline right here, I would highlight a few of those, and Phillip can give you more details. The most exciting one, I think, is the one that it's a Claudin 18.2 4-1BB bispecific antibody that we're developing as a bolt-on to the frontline gastric cancer, to bolt on to the standard of care, in this case, is IO and chemo. This agent is active in monotherapy studies and showed a very good profile in terms of safety as well as in combination, as well as efficacy when added to IO chemo. I'm happy to say that we are only weeks away from disclosing our escalation study data at ESMO GI on July 2nd. This is a major data release of the company, talking about the history of the company. I think this is an exciting milestone for us. Many of you may not think of gastric cancer in the light of being a significant disease in a large market like lung cancer or breast cancer, but it turns out there are significant medical needs. Globally, every year, we have about 1 million people diagnosed with gastric cancer, and 600,000 of those eventually die. The five-year survival is dismal at 7%. There is a significant medical need. That translates to significant commercial opportunities. If you look at the first line, Claudin-positive addressable market size is $12 billion, and Giva is positioned strongly to compete for that $12 billion opportunity. Finally, I would say I-MAB is in a very fortunate position that we have $168.8 million on balance sheet. That will fund us through some of the clinical readouts, for example, the escalation study that we're going to report at ESMO GI, as well as the expansion study that we're actively conducting right now. By the way, we're ahead of the schedule. We're very excited about that. We're going to share the expansion study first half of next year. Some of those studies will really drive additional growth for the company, and we're fortunate to have the resources to do that. Great. Now let's dive a little deeper into Giva. As you mentioned, Giva has shown very compelling monotherapy data in both efficacy and safety. Could you please walk us through the phase one monotherapy data, and what findings are you most excited about? I'll invite Phillip to give you the details. I think the excitement on Giva is, I want to take a step back, is related not only to the clinical data that I'll describe in a second, but also its molecular design and features of it as a molecule. Namely, it has very high avid binding to Claudin 18.2, about a log more avid than the leading asset in the class, zolbetuximab. If we look at the middle part of the molecule, the FC interacting portion, that's been mutated. We have engineered out processes such as ADCC and CDC that can be associated with on-target toxicity, again, demonstrated by the lead asset in the class, zolbetuximab. Finally, the 4-1BB interacting domains are scFv moieties, and they only interact with 4-1BB in the presence of Claudin 18.2, in the presence of a tumor cell expressing Claudin 18.2. What that translates to is that if our bispecific encounters a T cell in circulation, it doesn't stimulate it. We're actually seeing very little systemic immune responses. We're not seeing liver toxicity the way many 4-1BB molecules have been played out in the clinic. There are some molecular features that are distinctive. Our phase one study was notable for the fact that in a group of heavily pretreated gastric cancer patients, 45 patients now, and patients who had received the median prior three prior lines of therapy, we had a confirmed objective response rate of 18%. Notably, almost all of those patients had received prior IO. If you look at the toxicity profile, the grade three and above treatment-related adverse events were in the single digit, typically one patient out of 45. This is an incredibly well-tolerated monotherapy. When we looked at the development opportunities for Giva, we hypothesized that other Claudin 18.2 assets like ADCs or T cell engagers that more bluntly activate T cells through CD3 probably have a role to play in later lines of therapy, but they're going to be difficult to combine with the frontline standard of care. Giva, based on its phase one data, monotherapy data, really suggested that it was perfectly suited to be combined with the standard of care of immunochemotherapy. That's what we've done. On this very favorable safety profile, probably, could you please provide more context that this drug has lower dose reduction and lower dose discontinuation rate? Do you believe this very favorable safety profile could help translate tumor responses rate into more durable clinical response like PFS and OS in the first line? We have very low rates of discontinuation or treatment interruptions in the monotherapy data. I will mention parenthetically that a peer-reviewed paper describing the monotherapy data has been accepted for publication in Clinical Cancer Research. I encourage you to look there. This is data that I think will be very helpful and explain a lot of the field. The monotherapy data suggests that we did not need to interrupt. We did not need to discontinue patients. Again, if you look across the board, there are low-grade AEs. There is no reason to, again, interrupt or discontinue. What that means is that patients should be able to stay on study, on treatment in combination with frontline. What, again, we were very encouraged by, I want to emphasize this, the oral presentation at ESMO GI. This will be an oral presentation with review by an independent discussant. We're very excited to get that independent perspective as well. What you'll see is that in combination with nivolumab and FOLFOX, the toxicity profile looks very similar to that of nivolumab and FOLFOX alone, very similar to what was described in CheckMate 649 that led to the approval of nivolumab in that setting. Yeah, we are certainly really looking forward to the data presentation next month, I think, at ESMO GI, right? July 2nd. Yeah. Probably, could you take this opportunity to help investors set some expectation and probably what special data point, either efficacy or safety, that we should pay special attention to? This is the first, to Sean's point. I think we've been talking about presenting data, certainly as long as I've been with the company. This is the first volley in a long volley of data presentations at ESMO GI. ESMO GI is focused on 17 patients that were in the dose escalation part, part of our phase one dose escalation of Giva in combination with nivolumab and FOLFOX. Standard of care, nivolumab and FOLFOX, we don't modify the dose or the schedule. It's a bolt-on strategy. You will see two things. One, I just alluded to, which is tolerability. This is a very well-tolerated regimen with a toxicity profile that looks like nivolumab and FOLFOX alone. The efficacy is quite interesting. We have an extraordinarily high objective response rate if you compare it cross-trial for all those weaknesses that we can imagine and know about, a small study in phase one versus a phase three. Our objective response rate is far above that that was observed in either CheckMate 649 that led to the approval of nivolumab or Spotlight that led to the approval of zolbetuximab in combination with chemotherapy. We have a really impressive objective response rate. More importantly, there are anecdotes. There are patients in this study that I'll refer to as double negatives. What do I mean by that? I mean, these are patients that have PD-L1 less than 1%. These are patients that have low Claudin. In this small group of patients, it's a small group within a small study, we're actually seeing responders. Those patients would not be expected to respond to either nivolumab chemo alone or zolbetuximab chemo alone. We are seeing pockets and subgroups of activity that I think are exciting. It is a bit early for PFS. We do make an effort to show some PFS. We have follow-up of almost a year. We will present this data in detail in terms of the numbers. The PFS is still early. The patients are all, most of them are still on study. We have very few events. It is all very encouraging. To Sean's point, as we speak, we are continuing to complete enrollment in the dose expansion cohorts, the two highest doses that will be presented in July so that we can gain more knowledge in those subgroups and validate the results that we are seeing in escalation. Super excited. For the patient population with low Claudin 18.2 expression, we also noticed that during the monotherapy trial, you also show very encouraging data in this patient population with low Claudin 18.2 expression. Probably, could you please elaborate on why Giva could differentiate from competitors in this specific population? On the market opportunity perspective, what implications could this have for, I mean, expanding the market opportunity for Giva? Yeah, I think Phillip can talk about the science of why it's differentiated in terms of patient coverage. I will just say that this is to the core of why we believe Giva is strongly differentiated from competitors. If you think about the Claudin coverage, the only approved agent right now, zolbetuximab, requires a patient to have 75% of the tumor cell have 2+ 3+ stain intensity. For our clinical trial, we are looking for patients with 1% of the tumor cells as 1+ intensity. That significantly expands the scope in terms of Giva's patient coverage. There is yet another dimension, if you think about the 18.2 landscape. Zolbe is approved to be combined with chemo and chemo only, no IO components, which means when patients have a higher expression of PD-L1, their standard of care will not be Zolbe chemo. They will be IO chemo. In that space, Giva will add value by directly and simply combining with Nivo and the Chemo. When you take those two dimensions constraints into consideration, Giva's population coverage is an order of magnitude bigger than Zolbe, the first 18.2 approved asset. From that differentiated perspective, it is very strongly positioned, the Giva, to be competitive in the space. I'll turn it over to Phillip to talk about the science. Yeah, I want to, I view the distribution as well as science, right? Just to put numbers on it, we actually basically double the Claudin-positive patients that would qualify for Giva as opposed to zolbetuximab. Because we're seeing activity across the range of PD-L1, again, because PD-L1 above 1% is about 80%-90% of gastric cancer patients, those patients will be treated with an IO chemo backbone. Zolbe chemo alone is going to be limited use in those patients that are PD-L1 negative. It really is, we're vastly expanding the patients that can benefit from Giva. The hypothesis for why Giva works in a low Claudin setting, and to reference the monotherapy study, we had a responder with a Claudin level as low as 11% in the monotherapy study. I think it's due to the properties that I alluded to earlier, which is a very high affinity for Claudin. You don't need that many tumor cells. If you think about the Zolbe indication, it's not only 75% of the tumor cells. They have to express a lot of 18.2. It's 2-3+ staining, which is a surrogate for protein expression, right? They have most of the tumor expressing a lot of 18.2. Our threshold is 1% of tumor cells at 1+ or greater, so very faint or low protein expression. I think part of this is due to the fact that the affinity is very high for Claudin 18.2. As I mentioned, the precise activation of T cells by the fact that you have 18.2 in a tumor cell that binds to the part of Giva that interacts with 18.2, you have the 4-1BB interacting moiety on the other end. That leads to clustering of 4-1BB on the T cell surface and leads to activation of that T cell only because it is in the tumor microenvironment. Cool. During the past week, during ESMO GI, we see several data drops from different modalities targeting Claudin 18.2, I think including probably like two ADC and one CAR-T cell therapy, even in a small number of patients. Probably, how do you feel about this data? On a more broad perspective, compared to those emerging modalities, where do you see Giva standing out most clearly? How would you like to take advantage of Giva's unique feature moving forward? As an oncologist who, again, spent 20-plus years taking care of patients, I'm excited for patients to have opportunities if their tumors express Claudin 18.2 that they have options. I think that for modalities like antibody-drug conjugates, T cell engagers that really, again, bluntly and clearly activate T cells, sometimes in the tumor, sometimes outside the tumor, I view ADCs and direct T cell engagers through CD3 as being very good candidates for later lines of therapy. In other words, they have a higher objective response rate than Giva, but they come laden with toxicities. Toxicities, for example, grade 3 and above treatment-related adverse events as monotherapy, 60%. 60% of patients will have. That is more than really double what we see with Giva in combination, OK? There is a sense that these—I erred. I meant Giva in monotherapy. These assets have a future in later lines of therapy. Giva is perfectly suited to be combined with, again, the standard of care because you add it on. ADCs, I think, are highly unlikely to move into simply adding them on because they're too toxic. You can't take a regimen like nivolumab, FOLFOX that has a grade 3 and above adverse event rate of 55% and add another agent alone that has a 55% incidence of adverse events without taking something away from that standard of care because patients won't tolerate it, right? That's the difference. We hypothesized. We've now shown, as everyone will see on July 2nd, that Giva can be combined. Although, again, if I'm sitting in big pharma and I have an ADC or a T cell engager, I am thinking about how to move it in the front line. I am also very, very aware of the experience of other ADCs that you can't take an ADC, move it into an immunochemotherapy regimen without taking something away from that regimen. Super excited, really looking forward to the front line data on July the 2nd. Now let's move to another asset, PD-L1, 4-1BB, bispecific. Could you please introduce this asset and probably the developmental timeline and the key catalyst for this drug? Ragistomig is our PD-L1 4-1BB bispecific. It is being developed in collaboration with our partner, ABL. It is constructed very similarly to Giva in the sense that the FC moiety has been mutated, so there's no ADCC or CDC. The 4-1BB interacting part is very similar to Giva. The other end, where Claudin 18.2 is in Giva, that's PD-L1. I can summarize our phase I data by saying in a cohort of, we did multiple dose escalations in many different tumor types. In a highly pretreated population, again, median number of prior therapies of three, as monotherapy, we had an objective response rate of 27%. In these seven responders, five of which had received prior IO. In fact, one of the responders had a complete response that was a patient with ovarian cancer with seven prior lines of therapy. We had, I think, the properties of something that had requisite efficacy for monotherapy. When we looked, our line of sight for Raji has always been that we wanted to develop it in combination for IO refractory or resistant patients. When we looked at the rates of liver toxicity with Raji, we saw that the rates of grade 3 or above AST and ALT elevation signs of liver toxicity were in the 20%-25% range. We felt this was prohibitive. We have an asset that has activity as monotherapy. It has probably higher than we would want liver toxicity. We amended the phase one. We have new cohorts now recruiting in our phase one study where we're decreasing the dose and/or increasing the interval between cycles to allow T cells to recover. Because what happens when you continually stimulate T cells with 4-1BB, they can become exhausted. We saw no signs of this with Giva. I think the liver toxicity said to us that we should take another look at this. We are in the midst of doing this. We are in the midst of recruiting the first set of patients. I can tell you in the first set of patients, again, modifying the dose and/or schedule, we have not seen any grade three liver toxicity. We are very excited that we are going to be able to titrate the toxicity down, maintain the efficacy—it is very early for efficacy—and be able to really be positioned well to combine with other modalities, other IO. As this cohort plays out, we will also identify which tumor types appear to be the best to take forward in a phase two study. Thank you. For the post-IO setting, where already showed the compelling data, and during the past ASCO, we showed several data drops in these settings, such as PD-L1, ADC, and PD-1, IL-2 bispecific. How would you view this post-IO setting would evolve in the next few years and the potential market opportunity for I-MAB? Yeah, obviously, that's the holy grail. I spent many years of my career in big pharma developing checkpoint inhibitors. I will say that resistance to checkpoint inhibitors is the holy grail. There is a lot of activity in this space. I think the 4-1BB approach is a really interesting one. I think there are many studies that suggest that either through a bispecific approach like the IL-2, the modified IL-2 in combination with PD-1, that maintaining checkpoint blockade, a PD-1, PD-L1 checkpoint blockade with a novel IO agent makes sense. What we like is this single molecule does both, right, in a sense. Even if we wanted to maximize the checkpoint blockade, we could combine it with a traditional checkpoint inhibitor. I think we're clearly in a competitive field in the IO progressors, refractory resistant. We think that we have a very nice objective response rate. We had good sustained responses in those patients. If we address the liver toxicity issue, the rest of the toxicities were minimal. We have no CRS. We have no ICONs. We have no serious immune-related adverse events. Great. Probably let's spend a few minutes on the CD73 Monoclonal Antibody Uli. You previously guided that the developmental path for this asset really depends on the ongoing China trial. Maybe probably help us set expectations that what kind of efficacy and the safety data in the China trial would drive your decision to continue to develop this asset. I think the development of uliledlimab, or ULI, that's being done in partnership with TJ Bio is a really important development period for ULI for the simple reason that we've presented single-arm data combining ULI with a checkpoint inhibitor that suggested that if you select for patients that express CD73, it looks like you enrich the response for ULI plus a checkpoint inhibitor. The natural next step is to generate randomized prospective data. That's the critical study coming out of TJ Bio, which is a randomized prospective phase two study in frontline non-small cell lung cancer comparing ULI plus checkpoint inhibitor versus checkpoint inhibitor alone in a CD73 positive population. This study has completed recruitment. The analysis, they're waiting for events. We expect to have data available in the near future in the next several months. Obviously, if this data is positive, providing proof of concept that we can select patients based on CD73, this will be important information for us to consider as we consider further development of ULI. Cool. So finally, can you lay out the company's plan for the rest of 2025 and also 2026, like key milestones, catalysts, and that investors should pay attention to? Yeah, definitely. As you can tell from the Q&As up to this point, we got a few very exciting data flips coming in the near future. Most importantly, the Giva ESMO GI escalation data. This is the first combination data. You will see the position of the Giva in a very competitive and hot area. Last few days, we had the deal with Astellas, their license and ADC. That, to us, is wonderful news because 18.2 now not only is a commercially validated target, people continue to believe the value of this target and continue to believe in the value here. For us, that's a great news. The fact that the ADC, as Phillip talked about, is primarily a good candidate for second line and third line is not really a direct competitor, that particular asset, to us at this point. We think we are positioned strongly in the front line gastric cancer. ESMO GI is one data flip that's very important. We will continue to monitor that escalation data. When the data matures, we will communicate the PFS from that data. You will also recall that we have an expansion cohort for the two higher dose. Those two doses, those two cohorts are recruiting ahead of the schedule, reflecting the enthusiasm that we have for our PIs. This is an all-America study. All study sites are in America and the major cancer centers, Memorial, Mass General. The PIs look at the data. They have experience with monotherapy Giva, with escalation Giva. Many of them also have experience with Zolbe. When they have this side-by-side head-to-head comparison experience, in many cases, they put patients on our study, right? That's why we can recruit ahead of the schedule. To me, that is underscoring the value and potential of Giva in a professional setting. That gave us opportunity, coming back to your question, to continue to monitor data for the escalation, for the duration of the benefit. Coming into 2026, first half of 2026, we continue to guide that expansion data readout. We will communicate first half of 2026. For the other assets, for Raji, with the therapeutic window optimization phase one ongoing, we expect it to get readouts. It is an adaptive clinical design. It's going to take its course until we get to the end point. The expectation is that we're going to have a readout by end of the year. We may communicate that results early next year. For ULI, we are looking at a study that's currently conducted by TJ Bio. We don't have control of their timeline. Expectation is that second half of this year, we'll be able to see their results. That will inform our decision going forward. A lot of things going on, very exciting time for I-MAB. Very much looking forward to sharing more data as they become available. Great. We are right on time. Let's wrap up the session here. Thank you again for joining us today. Thank you for everyone to attend. Thank you. Thanks.