Good morning, everyone, and thank you for joining us at Needham & Company's 24th Annual Healthcare Conference. My name is Ethan Markowski, and I'm a member of the biotech research team here at Needham. Joining me today from I-Mab are Sean Fu, CEO; Joseph Skelton, CFO; and Phillip Dennis, CMO, who will provide a presentation on the company, followed by a Q&A session with the remaining time. I'll go ahead and turn it over to Sean. Good morning. Thanks for the opportunity. We're very excited to share the latest strategy and pipeline progress of I-Mab. All right, next slides, please. During the presentation, we will be making forward-looking statements, and the viewers should consult a forward-looking statement disclaimer for more details. Next, please. I'll be giving a high-level summary of the company's status before I invite Phillip Dennis to walk you through more details about the pipeline. I-Mab went through significant transformation in 2024, and that transformation continues, as you will see. Last year, we announced the divestiture from China assets. Next slides, please. Really changed the operating model and the strategy and the focus of the company from an old I-Mab, which at one point had over 200 employees, significantly broader pipeline and capabilities, including discovery, development, regulatory, even manufacturing. Very broad, very capital-heavy. To one that it is today, we are very focused. We are operating in the U.S. with only 24 full-time employees. All of them are onshore in the United States. We are focusing on three clinical stage assets, and all of them are clinically active. This is very different from what was before. By doing that, by focusing on clinical operations, by focusing on translational clinical development, we are much more capital-efficient, and we are guiding our cash runway into 2027. You will see that works really well when we think about the data flip in the next one to two years, some critical and exciting data readouts, and we have sufficient capital to see through those data flips. We are well-positioned to see the next phase of growth, and we've completed the transformation in 2024 through divestiture. Now we are focusing on the pipeline. Talking about the pipeline, let me give you an overview. The lead asset we have here is a Claudin18.2 4-1BB bispecific antibody. This is one we are developing as a bolt-on to standard of care in first-line gastric cancer, which is biochemo. This asset is active in monotherapy in heavily pretreated patients and showed good safety and exciting, promising efficacy signals in combination studies with IO chemo in front-line gastric cancer. Based on what we saw in the escalation study of 17 patients through three different doses, we are now conducting expansion studies for 40 patients in total across two doses to substantiate the value we saw in this asset through clinical studies with more data. This is a very exciting asset. We believe it's a best-in-class 18.2 4-1BB bispecific that has an active clinical development program in front-line gastric cancer. Very exciting. The second asset we have is a CD73 monoclonal antibody. We call it uliledlimab. This is another potential best-in-class asset. Through the divestiture, which I mentioned earlier, the China rights went to TJ Therapeutics, one of our business partners, and they had a license deal with Sanofi last year in which the greater China rights of this very molecule went to Sanofi. They did two things to this asset. One, it is a strong validation of the molecule, and we're excited about that. Second, it gave us a front-row seat to the clinical development programs that's ongoing in China that we believe will give us valuable input in guiding our next phase clinical development programs. Specifically, they're looking at uliledlimab, or Uli, plus IO versus IO in a randomized study in CD73 high populations. If readout positive, we believe this will significantly de-risk the asset, will give us green light to go forward full speed with our own development programs in the U.S. Third asset is a PD-L1 4-1BB bispecific. This one showed significant activities in monotherapy setting in heavily pretreated population. We are running clinical studies to optimize the therapeutic window, and we believe this asset has potential to help patients that suffer from solid tumors, and they are resistant to prior line of IO therapies. This is significant on my medical need, and we're very excited about the potential of this asset as well. You can see that we have a pipeline consisting of three clinically active assets. We're focused on advancing these pipelines going forward. With that, I wanted to turn it over to Phillip to give you more color around those exciting molecules. Phillip. Thanks, Sean. Next slide, please. I'm going to focus first on our lead program, givastomig, or Giva, and really emphasize that the elevation of Giva as our lead asset is a reflection of data that's emerged in the last several months and in no way reflects a lack of confidence in the other two assets that Sean gave an introduction to. Giva is a potential best-in-class 18.2 x 4-1BB bispecific. It has design features I'll go over in the next slide, but it has very high affinity for 18.2. It has clinical activity in our monotherapy phase I study. Again, I'll go over that in a second, in a very low level of 18.2 expression. This leads to localized T-cell activation in the tumor microenvironment that is intended to minimize 4-1BB-mediated liver toxicity as well as systemic immune responses. Our asset is the first to be tested in the United States in combination with immunochemotherapy as a standard of care in front-line gastric cancer. Next slide, please. This slide shows the mechanism of action. If we focus on the left, the bispecific is designed that if the bispecific through its 4-1BB domain encounters a T-cell in the circulation because of a lack of clustering due to the fact that there's no tumor cell there, that in fact, that T-cell will not become activated. In contrast, on the right side of that diagram on the left, if you see a tumor antigen 18.2, you see several molecules of givastomig interacting with the tumor cell, interacting with 4-1BB on a localized T-cell. You can see that that leads to T-cell activation. Moreover, if we get into some details of the molecule, again, if you look at the 18.2 binding domain, it has a much higher affinity, a log higher affinity in terms of for 18.2 than the lead asset in the class, Zolbetuximab. It has a silenced Fc portion, so there's no ADC or CDC. This minimizes unintended systemic immune activation. Finally, as a 4-1BB agonist, the scFv domains at the bottom there lead to localized T-cell activation in the microenvironment, leading to potent tumor cell killing and minimal liver toxicity and systemic immune response. Next slide, please. This is a comparison. We know that the 4-1BB, sorry, the Claudin18.2 class is very crowded, especially in gastric cancer. What we've done here is to try to highlight some of the mechanisms of action of the leading competitors and to show how Giva compares very favorably. The Giva mechanism of action that I just described is at the top with number one there. There are T-cell engagers such as the Innovent T-cell engager, and these typically engage, especially if they use CD3. This induces tumor cell killing by T-cell-dependent cellular cytotoxicity and bystander killing via activation of the T-cell receptor co-receptor, CD3. This leads to typically a very high rate of cytokine release syndrome, and I'm going to show that data again in a second. Importantly, there are other assets such as antibody-drug conjugates, such as from AstraZeneca, AZD0901, that leads to killing by cytotoxicity due to delivery of the payload and bystander killing, inducing ADCP and CDC. Here, there's a high risk of on-target GI toxicity and off-target systemic toxicity, especially with premature release of the payload. Finally, the lead asset in the class, again, Zolbetuximab or Zolbi, kills cells by ADC and CDC, and it has a very high risk of on-target GI toxicity. Next slide, please. This is a snapshot of our phase I data presented last year. The swimmer plot is shown on the left. Without getting into too much detail, what we can see is that of the responders in this graph, there are several long-lived responses in heavily pretreated patients. In fact, that patient in orange, about the seventh one down, has been on study for about two years at this point. Why is this important? This is important because these patients had received a median prior lines, three prior lines of therapy. This represents 43 Claudin18.2 positive patients. Importantly, we use a cutoff of 1% of cells at 1 plus or greater intensity. Of these responses, we observed seven of them for an ORR of 16%. The responders range from Claudin18.2 expression of 100 all the way down to 11%. In addition, of those seven responders, five had received prior IO. Next slide, please. In addition to the efficacy, the tolerability was quite notable. We did not encounter a dose-limiting toxicity, and we did not reach an MTD. The chart to the left shows that in terms of treatment-related adverse events, there were virtually no grade four or five. The one exception is a grade four thrombocytopenia that was not associated with bleeding, arguably suggesting that this is not clinically meaningful. Moreover, if we look at the rates of clinically meaningful adverse events such as nausea and vomiting, we are at 2%, one out of 43 patients at grade three. This has really suggested that Giva has a very special place in terms of having monotherapy activity in a very highly pretreated population with gastric cancer. Moreover, it was very well tolerated as monotherapy. Next slide, please. In fact, if we compare Giva against the early phase I data of Zolbetuximab, which is again a naked antibody, we can look at the first row that shows the cutoff for 18.2. Again, ours is the minimal really that one can think of, which is 1% of cells, one plus or greater. At the same cutoff, Zolbetuximab had a 0% ORR, 0 out of 15 patients. When they tightened the threshold to 50% of cells, two plus or greater, they had an ORR of 9%, suggesting again that Giva had greater efficacy in the same group of patients, although it is a cross-trial comparison. Moreover, if we look at important toxicities such as nausea and vomiting, we have a much better safety profile, which again will be shown in just a second. Next slide, please. This is a comparison of givastomig on the left, zolbetuximab, the second to the left, and then the lead ADC, arguably, CMG901, and the lead T-cell engager from Innovent IBI389. This slide is very detailed, and I already went over the differential mechanisms of action. Importantly, this is data from the monotherapy phase I data. You can see that the threshold is very different for these assets, right? With us being the broadest inclusion criteria, arguably we offer the greatest benefit to most patients. We can see that the ORR is below that, but importantly, the safety profile is what I want to talk about, which is in fact the, if we look at treatment-related adverse events, I've highlighted this in red. What this shows is that if we look across the board at grade three or greater treatment-related adverse events, overall Givastomig led to a 33% incidence. This was not reported in Zolbetuximab, but for the ADC and T-cell engager, we're talking almost double the rates of grade three or greater adverse events that were treatment-related. If we look at other important adverse events such as nausea and vomiting, we can see again how we are several fold less than Zolbetuximab in terms of grade three or greater nausea and vomiting. Importantly, we included CRS in here because of the T-cell engager. Again, although the majority of them were grade one or grade two, each of these grades leads to patients being admitted into the hospital overnight. This is an important distinguishing fact. In fact, if we look at the profiles of these assets, and as someone who used to lead a major ADC program at a big pharma, I can tell you that the T-cell engager and ADCs do have a profile that's quite typical and quite favorable for development in second and third line settings. However, one can look at their profile and say they are not well suited to be combined with standard of care for gastric cancer, which is IO plus chemotherapy. In contrast, givastomig has a really interesting profile, well tolerated with notable monotherapy activity. That is what we tested. The next slide, please, will show our phase I design. This is a phase I study designed, and in fact, the blue part is completed now with 17 patients where we simply added givastomig at three different doses to standard of care nivolumab and FOLFOX. This is in front-line unresectable patients, HER2 negative, with again a minimal level of 18.2 expression at 1%. In the dose escalation, we observed no dose-limiting toxicity. We did not reach a maximum tolerated dose because we did not dose escalate beyond the point at which our biomarker plateaued. When we looked at the efficacy, what we noticed, especially in dose level two and three, is that almost all of the patients had a response. In fact, if you total the objective response rates for this cohort and compare across study, at the two studies that led to the approval of Nivolumab plus chemotherapy or Zolbetuximab and chemotherapy, our objective response rates were far in excess of those numbers reported in those phase three studies. This led us to look at the plans for dose expansion and to amend the study to include more patients because we were so impressed with this data that we really wanted to validate it. In fact, our recruitment for the dose expansion part shown on the right, dose expansion part shown on the right, we are far ahead of schedule. We're months ahead of schedule in terms of recruitment. To date, when we look at the early data coming in from dose level one in terms of the expansion cohort, the objective response rate is in fact being corroborated in the expansion cohort that was first established in the dose escalation cohort. We are very encouraged by this data. In fact, the dose escalation data will be presented at a major meeting in the second half of this year, and the dose expansion data is planned to be presented at a meeting in the first half of next year. Next slide, please. This slide shows both the market opportunity and really the clinical opportunity shown on the right. In terms of the market opportunity of the 250,000 cases of gastric cancer, about 137,000 are 18.2 positive. It is a very large market. If you look at the place that Zolbetuximab occupies, given its indication with chemotherapy only, this rectangle is drawn to scale. In fact, it represents about 8% of the total gastric cancer market. That is because of the fact that Zolbi's indication is at 75% or greater Claudin18.2 expression. Because its indication is with chemotherapy alone, the use of that is really predominantly in the PD-L1 less than 1% that is shown on the y-axis. This leaves a lot of space for Giva. Giva is potentially first in class in that purple area because the odds of Zolbi ever getting less than 75% are quite low because when the Claudin18.2 expression falls off, the activity of Zolbetuximab drops precipitously. Even if Zolbetuximab moves into the IO plus chemo space shown in blue above the Zolbi rectangle, what this shows is in fact that Giva still has a great chance to be best in class there because, again, in the phase I monotherapy data, Giva had better efficacy and was better tolerated. We think the potential for Giva is quite the upside is tremendous here. The opportunity does not exist in gastric cancer. We have not even listed other stages of gastric cancer, such as locally advanced but resectable, but in fact, other tumor types as well. Next slide, please. If we establish proof of concept, which it certainly looks like we're on that track with gastric cancer, by combining Giva with IO and chemotherapy, why not move into other GI malignancies where Claudin18.2 expression is quite prevalent, especially in tumor types with high unmet medical needs, such as pancreatic cancer and cholangiocarcinoma shown on this slide. The market opportunity is large here with $6 billion and $3 billion respectively for each of these tumor types. Again, most of these tumors express Claudin18.2. For pancreatic cancer, the standard of care is a chemotherapy doublet. Why not combine Giva with the chemotherapy doublet if we know that we can add it to IO and chemo safely and improve efficacy? The standard of care in cholangiocarcinoma is in fact IO plus chemo, so why not do that experiment as well? The upside here is tremendous. Now, the development of Giva, next slide, please, is also secured by intellectual protection, and that's just listed here in terms of the patents in the U.S. that have been granted for composition of matter and method of use that extend protection to 2040. Importantly, there are other patents that are pending, both in the U.S. and globally, related to method of use and the combinations that extend the protection through 2043 and 2045. Next slide, please. In summary, Giva is a very high affinity 18.2 bispecific with conditional 4-1BB activation. I didn't point out, but in fact, we have again minimal liver toxicity, no cytokine release syndrome, so we're getting really controlled activation of T-cells in the microenvironment. It is designed for long-lasting immune responses across a wide range of Claudin18.2 expression, and our response rate and tolerability stand out amongst other assets in the class that really points out the potential to combine givastomig with an IO chemo backbone in front-line therapy. We have important catalysts with our escalation data being presented in the second half of this year and the dose expansion data being presented in the first half of next year. Next slide, please. I will touch on uliledlimab. In the interest of time, I'll be a little bit brief here. Uliledlimab is arguably the best CD73 antibody out there, and I will show reasons why in a second. CD73 is the best target in the adenosine pathway. Why target adenosine? Because it's an immunosuppressive metabolite. In fact, uliledlimab, especially compared to the lead antibody in the class, oleclumab, that's being developed by AstraZeneca, we have better in vitro inhibition of CD73. As Sean mentioned, we are awaiting data to emerge from our collaborator in China, TJ Therapeutics, that is doing a prospective randomized study in CD73-selected patients of uliledlimab plus IO in front-line non-small cell lung cancer. Our development is being paused until this data emerges. Next slide, I want to point out that really the characteristics that make us excited about this asset. In the interest of time, there are several ways to target the adenosine pathway. CD73 is arguably the best point to inhibit the pathway because it is a point of convergence. There's no bypass around it. It is the only enzyme that converts AMP into adenosine, and it is the rate-limiting step. Next slide, please. Now, the advantages of uliledlimab compared to oleclumab have to do with how uliledlimab interacts with CD73, which is a dimer. Uliledlimab acts in a way to form an intradimer binding mode, which leads to dose-dependent inhibition, as shown on the upper right. In contrast, oleclumab forms interdimer binding modes, which leads to this concept called the hook effect. I can point that out to you in the next slide, what that looks like if you do an enzymatic assay. Next slide, please. On the left is the ability of uliledlimab to inhibit CD73 enzymatic activity in an in vitro experiment. Different doses of uliledlimab are shown in blue. As you dose escalate, you can see that the% of enzymatic activity falls to zero, complete inhibition of CD73. In contrast, when oleclumab is dose escalated at the same dose ranges, what you can see is that there is inhibition initially, but then the inhibition is paradoxically lost as you increase oleclumab concentrations. This is what's called the hook effect. I'll also point out that oleclumab never leads to complete CD73 inhibition, again, suggesting that uliledlimab is a better inhibitor of CD73. Next slide, please. Does a hook effect exist in a model system? The answer is yes. Next slide, please. This slide shows that, whoops, I'm sorry, too fast. I apologize. This simply shows that if you, in a mouse model, if you dose escalate uliledlimab shown on the left in the shades of blue or oleclumab shown on the right, then in a dose-dependent manner, you get dose-dependent inhibition of CD73 as well as dose-dependent tumor inhibition. In contrast with oleclumab, you get neither. Next slide, please. This is the clinical data, single-arm data that showed that really got us excited about our concept, which is really differentiated in the field of CD73 assets, which is we think we can identify patients who will respond better to U li by measure CD73. That's shown very simply in this slide, which shows that in a front-line non-small cell lung cancer cohort, if you divide the patients by CD73 and PD-L1, this slide shows that for CD73 high, the ORR to Uli plus a checkpoint inhibitor that's used in China, toripalimab, that the ORR is approximately twice that of the CD73 low patients in the row below. In fact, if you look across trial at KEYNOTE-0 42, which is a landmark study with pembrolizumab, the ORR in that study in the same PD-L1 group of greater than or equal to 1% was 27% versus 63% in the CD73 high group. Again, that was being tested in combination with Yuli. I will say the obvious that in fact Yuli and Tori were very well tolerated, so we anxiously await the development of the data from TJ Bio with their randomized prospective data. Next slide, please. Let's go to the slide on Ragi, and I'm conscious of time. Ragi is a really interesting 4-1BB bispecific. It leads to conditional activation of T-cells whenever there's a tumor cell that expresses PD-L1, and in fact this is the implication here to mitigate liver toxicity as well as systemic immune response. Next slide, please. I want to point out some of the phase I data, which is really interesting. This is phase I data that was presented at ASCO last year, 44 evaluable patients, and again, most of them had at least three prior lines of therapy. If we look at data at the three and five milligram per kilogram cohort, which we think are really the minimally clinically active doses, we saw an objective response rate of 27%. This included one complete response and six confirmed partial responses. The complete response was observed in an ovarian cancer patient who had received seven prior lines of therapy, and almost three quarters of them had received prior PD-L1 inhibitors. This is really exciting to us, and I think that this is very exciting as monotherapy data, right, in IO refractory patients. Next slide, please. What we encountered was that we did encounter an MTD at 7 mg per kg every Q week, but our most common treatment-related adverse events were increased AST and ALT, which are signs of liver toxicity, which is quite common in the 4-1BB class. In fact, if we look, next slide, please, if we look at the comparison with acasunlimab being developed by Genmab in their phase I data, it is clear that we have a far superior ORR at 27% compared to 7% or 13% depending on which cohort you look at for acasunlimab. Importantly, their AST and ALT rate of grade three or greater was at 10% compared to ours at 25%. This has led us in our development of Raji and collaboration with ABL Bio, what we're doing is actually amending the ongoing phase I to test lower doses and less frequent doses because we think by doing that, we will actually increase the therapeutic index, will minimize T-cell exhaustion, and we eagerly await that data. Next slide, please. In summary of our key readouts, I'll mention we have a financial runway into 2027 to complete the trials that I've mentioned to you today. We have important readouts for Giva in the second half of this year for the dose escalation data in combination of front-line gastric cancer. We have the dose expansion data that will be presented at a major meeting in the first half of next year. The prospective data of uliledlimab should be available in 2026, and we expect the ragistomig data with the different dose and schedule cohorts to be available by the end of this year. With that, I'll stop and take any questions. Great. Great. Thank you, guys, for the presentation. At this time, anyone attending the call is welcome to post their question using the question and answer feature, and I will be keeping an eye on that. Maybe a couple from me to start. Kind of a broader question, it's a very hot topic right now, which would be tariffs. I know in the very beginning, you did a nice job of illustrating the changes of I-Mab structure in the recent years and divestment out of China, but wanted to know if there is any impact or expected impact from tariffs, or do you think that's largely mitigated with I-Mab is currently set up? I think you're muted, potentially. Hope this is better. Thank you for the question. Yes. Yeah, I think the divestiture was a strategic move that proved to be smart to get ahead of the curve, so to speak, and to avoid any of the potential issues from the geopolitical tension as well as the tariff, which is a more recent topic. For us, I don't see the tariff going to have an immediate impact. Our escalation study and expansion study for givastomig, our lead assets, are conducted in the U.S., and all the data that Phillip mentioned were U.S.-based studies. Our supplies of clinical materials are in place for the next phase of clinical development. Overall, I don't see tariff being a significant impact going forward. That That positions us, and we have enough cash to see through some of those data flips. I think we're positioned in an enviable position with the pipeline being active and differentiated and with the resources to see through. I think we can really focus on execution without having to worry about outside exogenous variables that we don't have immediate control. Great. No, thank you for that input. Another question that we get quite frequently, and you did touch on in the beginning, is kind of if you could maybe walk us through a little more the decision-making to prioritize Givastomig. I know you hit on it, but is it mostly the data, competitive reasons, combination of different variables? Any thoughts you have there? No, I think those are really the important points feeding into our consideration. It really is the data coming out of Giva, and it's a unique position of this molecule amongst competitors. If you look at the holistic profile of the molecule, Giva threaded needle with a very nice combination of pronounced efficacy and very benign safety profile, especially in combination with standard care without changing any of the components of the standard care. That makes the clinical practice and the later phase clinical development that much straightforward, and it's favored by regulatory agencies if you don't change the standard care. That is a unique profile when you look at other modalities, be it ADC or bispecific CD3 engagers or monoclonal antibodies. The ability to combine with standard care unchanged, it is quite an advantage for Giva. Now, with Yuli, I think it's still a very strong molecule and potentially best in class. What we are doing is not to say that Yuli is going to be delayed. In fact, we're going to say, "Hey, we're going to look at this molecule in a capital-efficient way." There is an ongoing clinical study that will significantly de-risk our next phase development. Why don't we give it some time for data to mature before we put down our own capital to develop the asset in the U.S.? It's a free option that we wanted to take advantage of. It's the exact same molecule. I think that's the right thing to do. We're looking at a data readout from that study sometime in 2026, and we're comfortable with that timeline. To add to that, we have an FDA-approved IND study plan in our hands. Once that data comes back positive, we're ready to go. Sean, if I could just add, again, we all are enthusiastic about the pipeline as well as especially Yuli as a best-in-class CD73 antibody. I want to go back to Giva. One thing I did not mention that I intended to is that it's based on data, the emerging data. The data that I mentioned, it's not just the objective response rate that is far in excess of what's been reported in the phase three studies that I mentioned. It's the fact that patients, we're seeing responders in the dose escalation study that are PD-L1 negative or Claudin18.2 at 1%, bare minimums. These patients would not be expected to respond to either Nivo FOLFOX alone or Zolbetuximab FOLFOX alone in the study called Spotlight. This was really part of the exciting data that emerged. I just want to reinforce that, that it's not only the data in 17 patients with the overall objective response rate, it's the fact within that response rate, there are really outliers, anecdotes that really suggest that there's something going on here that's worth, that's why we're exploring it more fully in the dose expansion cohorts. Great. Great. Thank you for those responses. Maybe highlighting or double-clicking a little on the Giva safety data, you showed a nice chart, kind of a cross-study comparison and how it appeared better tolerated than some of the other lead programs. I was wondering, yeah, because I didn't see it, I was wondering if these reduced safety events translated into lower discontinuation rates at all or if those were pretty comparable. Just if you remember. Yeah, the discontinuation rates, we don't have that on the slide. The easiest thing to state is that they were comparable. Again, if you think about it, the idea here is not, we're not trying to make the case that Giva's a better monotherapy than an ADC or T-cell engager. It's a better monotherapy to be combined with front-line therapy, right? When you take any of these assets that you're looking at in the slide and you combine it with IO and chemo that already has a very high rate of treatment-related adverse events, you have to be very careful with what you add, especially if you're not going to modify the dose or schedule. To me, the discontinuation rates and the monotherapy are, again, it's slightly better with Giva, but that's not really the message. The message is that this improved toxicity profile translates into the ability to be combined with other agents. No, that makes a lot of sense. I am monitoring questions from chat, but maybe one more for me in the meantime. I know at I-Mab, business development has always been a consideration, especially maybe for a later stage asset. I have talked of in-licensing a program in the past. Is that something that is still open or in focus or less of a priority now and more focused on really clinical development of Giva at the moment? We're razor-focused on the execution of the current pipeline. We see value in every molecule, especially givastomig, our lead assets. We're putting our resources and our energies behind this molecule to advance it as fast as possible. Having said that, we have our ears on the ground when thinking about the growth potential beyond the current three pipeline assets. We are paying attention. We're having conversations. We're looking into future growth opportunities, but that is complementing the execution efforts behind those three assets. These are our priorities. Great. Great. I'm taking a look, and I don't see any further questions from the chat at this time. Just want to thank three of you for coming on and talking a little bit more about I-Mab. Thank you for attending our conference. Thank you very much, Eitan. Thank you.