Oncology Innovation Summit. I'm Philip Nadeau, one of the biotech analysts here at TD Cowen, and it's my pleasure to moderate a fireside chat with SELLAS Life Sciences Group this morning. We have with us today Angelos Stergiou, who is the President and Chief Executive Officer of SELLAS. Angelos, I'll hand it to you first. Could you provide a brief state of the company? What are the most important events for the company over the next year, and what does SELLAS need to do to create shareholder value? Firstly, thanks for having us, Phil. We are a late-stage oncology company with focus on the development of galinpepimut-S or GPS, which is an innovative, potentially first and best-in-class Wilms' tumor 1 or WT1 targeting heteroclitic immunotherapy, which we licensed from Memorial Sloan Kettering Cancer Center here in New York. It is currently in a phase III study, AML study. We also have a highly selective and specific CDK9 inhibitor, SLS009, which is in phase II, and we're really currently focusing on Acute Myeloid Leukemia with both of our assets. Quite exciting, we may be able to really shift the AML treatment paradigm, among others, from induction all the way through maintenance with SLS009 and GPS, respectively. As for what shareholders will be looking out for, I think the key catalyst, as for GPS, as I mentioned, we're in a pivotal phase III study, the REGAL trial, and that is in patients after second complete remission. We expect data here very soon. The study's based on work that we have done previously in CR1 and CR2 settings. We recently announced the 78th out of 80 events or deaths in the study. The 80th event is the trigger point towards analyzing the data and then disclosing top-line data. I think both the achievement of the 80th event followed by top-line data will be announced as soon as those events materialize, and I think are key catalysts for SELLAS. As for SLS009, our selective CDK9 inhibitor, we're currently in a frontline AML study. Really utilizing our extensive transcriptomics, genomics, and proteomics work that we have done, really down the path of precision medicine to identify patients most likely to respond with our treatment. This is on the heels of exciting data that we had seen in the previous relapse refractory AML study. Here we expect to have top-line data available from the study later this year, both response rates and survival outcomes. That's really helpful. Maybe we'll dive into GPS in a little bit more detail. Can you talk about the rationale for WT1 as a target in AML? In particular, how widely is it expressed, and are there patient subgroups on which it's most prominent? Yeah. Our immunotherapy, galinpepimut-S or GPS, it targets the number one-ranked cancer antigen by the National Cancer Institute, Wilms' tumor 1 or WT1. WT1 in the fetal state, it's responsible for kidney formation, among other functions it has. Once we're born, it disappears and then comes back again when it causes cancer. It's a fetal oncoprotein, functions as a nuclear transcription factor, and it's overexpressed in about 20 tumor types. AML is our lead indication, and more than 97% of AML patients express WT1. We've also done work in other indications, with positive data like mesothelioma, where again, WT1 is expressed actually in all patients, as well as in multiple myeloma and ovarian cancer, with expressions ranging up to 50% of patients. Can you talk a little bit more about the structure of GPS? How does it spur an immune response? Yeah. GPS is a mixture of four synthetic peptide sequences derived from the WT1 protein. The peptides range from nine to 22 amino acids in length, and we truly have a differentiated immunotherapy with GPS, as it was precisely and deliberately designed to select the right cancer target, WT1, and it's a mixture of four peptides, which are fragments derived from the WT1 whole length protein. We have utilized heteroclitic concept in two of the four peptides. It's a by-design engineered mutation of amino acid sequences to further enhance immunogenicity. That allows us also to treat patients over a long period of time through administration of booster inoculation. Really, this substitution alone leads to a 70-fold stronger binding of the peptides to the MHC class I pocket. As it relates to how the immune response is really evoked, because our multivalent approach targets 25 carefully selected WT1 epitopes, and we have a really high prediction of optimal MHC class I and II presentation to be really maximally immunogenic both for CD4 and CD8. GPS works through the induction and stimulation of T lymphocytes, both cytotoxic CD8 cells, which are attacking the tumor directly, and CD4 cells, which are very important for immunologic memory and maintenance and helper function of the cellular immune response. Again, two of the peptides within the GPS mixture are native. The other two, we have the by-design modification by amino acid sequences. In that sense, the heteroclitic WT1 peptide, it carries a mutation, and is presented to the native CD8 cell through the host antigen-presenting cells, macrophages, dendritic cells, monocytes. The CD8 cells then gets activated, becomes a cytotoxic T cell, which now is able to recognize not only the mutated version of the WT1 peptide against which the host was immunized, but also the corresponding native WT1 fragment. The native fragment then gets expressed and gets presented on the membrane of the cancer cells in an MHC class I context, and similar events happen also with CD4 cells with cross-presentation of the WT1 heteroclitic fragments. How does that translate into benefit in AML? What's the proposed mechanism of action in AML? Yeah, I think for AML, as we have seen before in all our work, you should think of it basically as an immunosurveillance. When a patient is in remission and we target as a monotherapy indications because of the tumor microenvironment where patients are in CR, in complete remission. Basically, through the inoculation of GPS, we basically provide the immune system a picture of what the cancerous cells look like. If and when the cancer comes back, we can then really laser focusly combat and kill the cancers, which ultimately translates to prolongation of survival. Can you review the data that GPS has produced as a post-CR maintenance therapy in AML so far? Yeah, I think there are three key programs or studies that we've performed both in CR1 and CR2. In the phase II AML trial, in patients after first complete remission or CR1, which included 22 patients and the majority had unfavorable cytogenetics. We saw a median overall survival of close to 68 months, with a median follow-up of around 45 months. This activity is quite unprecedented, and the study was deemed positive for the primary endpoint. According to the pre-specified efficacy threshold, it was deemed positive because the protocol had said, if we see at least a 34% survival rate at three years, it would be deemed positive. We actually saw 50% survival. Focusing on patients 60 years and older in the phase II trial, which inevitably make up the majority of AML patients, we saw an unparalleled median survival of 35.5 months, which corresponded nicely to the phase I work that we had done, where we saw a 32.3-month median overall survival. In addition to that, with strong immune responses in excess of 80%. In the phase II study, in the CR2 setting, which is the setting that we're currently in the pivotal phase III trial, those were patients who had already undergone frontline therapy. They achieved CR1, they were subsequently followed without undergoing a stem cell transplant, eventually they relapsed, and they underwent another round of second-line therapy to achieve second complete remission. Here again, was an opportune setting for us with our GPS drug. The median survival in that patient cohort was 74 years old. We showed a 21-month versus 5.4-month survival benefit in favor of GPS with a P value of 0.02. I think what is key here, Phil, is really with our GPS treatment in AML as a monotherapy, that patients are in complete remission because of, as I mentioned, because of the tumor microenvironment. Could you outline the design of the ongoing REGAL pivotal trial? Yeah. The REGAL study has enrolled 126 patients, one-one randomized. Roughly half of the patients received best available treatment, which is clinician's choice. It can be venetoclax, hypomethylating agents, or observation, versus the other half of patients on GPS, which is administered every other week for about 2.5 months, and then followed by monthly and quarterly booster injections thereafter. The last patient was enrolled in March, April 2024. What were the specific inclusion/exclusion criteria? Yeah, I think the key inclusion/exclusion criteria, those were AML patients not eligible for transplant and targeted therapies less than six months after response was observed, and patients had to have at least, if I recall, 300 lymphocytes per microliter to maximize the GPS effect, because GPS is an immunotherapy whose effect is mediated by lymphocytes. Can you describe the regimens in the treatment and control arms of the trial? Yeah. In the treatment arm, patients receive only GPS as monotherapy. GPS arm patients, they receive GPS, I mentioned, every two weeks for 10 weeks. After that, they receive monthly and quarterly booster injections if they remain in CR. In the control arm, patients receive physician's choice, best available therapy. Again, venetoclax, hypomethylating agents, low-dose cytarabine, or observation, or a combination thereof. Again, it depends on the physician and how well the patient can tolerate any of those treatments. There's been some modifications to the dosing in the treatment arm over time. Can you describe those changes and why they were made? Yeah. In the GPS arm, we have extended the duration of treatment from one year to two years, and then to three years, and then beyond three years. That was done at the request of physicians who assessed that the patients benefited from GPS therapy beyond one to two and three years or longer. There were no changes made to the control arm regimens, and that's the only piece of information we have as a sponsor, as we're blinded to study outcomes. What have you disclosed about the powering of this study for its primary endpoint? The survival hazard ratio is the primary endpoint, and hazard ratio, as you know, compares proportion of patients alive at multiple time points in one arm versus the other by calculating probability of survival between the arms at all time points. Hazard ratio of 0.636 means that overall, over time, patients in one arm had probability of dying of 63.6% compared to the other arm. In other words, patients in one group have 36.4% lower probability of dying than patients in the other group at all time points. The final analysis at 80 events has been designed, or the program has been designed to provide at least 90% power under an assumed hazard ratio of 0.636, and that's an estimated, as an example, and it corresponds to eight months on control versus 12.6 months on GPS. You mentioned 78 events have occurred. We're waiting for the 80th event, presumably it'll happen sometime soon. At the time that enrollment was completed, when was the 80th event expected to occur? In the initial assumptions, the 80th event, which is a trigger point for analysis, was expected to occur approximately 12-15 months after the last subject was randomized. However, 80 events have not occurred yet, almost 25 months after enrollment completion, which obviously indicates longer survival than projected. Following the announcement of the 80th event, when do you think top-line results would be available? Also, what do you expect to include in that top-line release? What measures will you disclose? Yeah. Well, we will provide further guidance around timing of top-line data when we announce the 80th event. We will certainly provide the key data around the primary endpoint, hazard ratio, survival, and so forth. You mentioned the powering the study is for 0.636 HR. What median survival do you expect in the control arm? Yeah. The control arm, historically speaking, survival with low-dose cytarabine or hypomethylating agents was six-nine months. With venetoclax added to hypomethylating agent, there's some increase to 10 or 11 months in trials. Those numbers also include transplanted patients, it's not easy to assess for patients in our trial, which does not include transplanted patients. I think what's important here, Phil, is that clinically, what is important and what we would want to see in such a setting is anything, because it's a relapsed and then patients in remission, conservatively, anything more than a four-month difference between the two arms would be deemed very clinically significant. Obviously, we're very hopeful to have even a bigger difference. I think the key question investors are asking is that it's been over two years since the completion of enrollment, so events are definitely accruing much more slowly than anticipated. Aside from GPS prolonging survival, are there other plausible reasons that it's taken so long for events to accrue? I mean, it seems to bode well for a trial in which the survival curves, in an indication where the survival curves are pretty well understood, seems to bode well that the trial's taking much longer to complete than expected. Yeah. I think for those of us who live in the immunotherapy world, I think it's sort of a known phenomenon of the separation of the curves taking a bit longer and down to the tail. Survival times in REGAL, indeed, as you mentioned, appear longer than expected from a mathematical standpoint and current treatment landscape, again, published data and real-world evidence. The fact that we administer GPS to patients even three years after enrollment, every passing month, again, I think increased probability of a successful study due to potential GPS effect, as we have seen in the previous CR1 and CR2 work. We had also randomly picked several GPS patients to assess immune response, which was similar to the prior CR1 and CR2 work. Also, although toxicities are commonly observed with therapies in the control arm, physicians are allowed, obviously, to modify them. We have not seen any on the GPS arm. Other considerations could be that patients are followed more frequently in the phase III study versus ordinary patients, which are followed up in real-world settings. Infections, for example, can be treated more effectively and timely. Really, it's difficult. I would not want to speculate of any specific reason why patients are living longer in the REGAL study. I think for reasons I just mentioned, we hope and we believe that GPS is really the driving factor here as it relates to patients living longer. Can you provide an overview of the AML market, and in particular, how many patients are in CR2 each year, and what percent of those would be eligible for therapy? Yeah. The incidence rate of AML is around 85,000 or so in the U.S., E.U., Japan, China. In the U.S., there are approximately, depending on the data, between 22,000-27,000 patients annually diagnosed with AML. Around 60% or so of patients will achieve complete response, depending on a myriad of factors. Due to the severity of the disease, the majority of the patients who did not receive a stem cell transplant will ultimately relapse. Of those relapse refractory patients, about 40%-50% will achieve a second complete remission. If you bring it down, it boils down to about 4,000 patients per year. Certainly, with the advent of new treatments and SLS009 being one of them, and targeted therapies, more patients get into remission, which really broadens the market size for us for GPS. Are there any other therapies used for maintenance in CR2 patients today? No, there's nothing approved in the AML CR2 maintenance setting. GPS, being an effective immunotherapeutic with a favorable safety profile, I think that's a very important nuance that even we forget to mention a lot of times because those patients have really seen myelosuppressive, really toxic drugs. We have a favorable safety profile. It would really indeed set the stage, I think, to possibly become the maintenance therapy in the AML after patients are in remission, and not just in CR2, but potentially also in the CR1 setting. Based on the data that we've previously reported, I think there's a great interest also in evaluating GPS in combination with bone marrow transplant in AML. I think lastly, I would also expect, depending on the data that we will see, and what I believe, is that there's a good chance that the REGAL trial will show GPS to have a medium survival, very similar to what Onureg showed in AML CR1 to get NCCN guidelines also in the CR1 setting. We're very bullish, but at the same time, we're very pragmatic. Great. Maybe moving on to SLS009, can you provide a brief overview of that program? In particular, how is SLS009 differentiated from other CDK9 inhibitors that have been developed in the past? Yeah. CDK9, as you know, is a major cancer target. Briefly, it's well known that many cancers depend on rapid production of proteins to promote cancer growth, like MYC and then also, once they're damaged to anti-cancer drugs like MCL1. To really rapidly forge those proteins, cancer cells crucially depend on the activity of certain factors, among which is CDK9, that allows elongation of RNA transcript necessary for the production of proteins that keep cancer cells alive. By shutting down CDK9, SLS009 blocks the production of cancer-promoting and protecting proteins, which results in the death of cancer cells. Other programs, such as AstraZeneca, they are CDK9 inhibitors who discontinued their programs. I think in part it was they're not as selective and specific to the CDK9 target as SLS009 is, and we have seen that through experiments we have done looking at other kinome proteins. Basically, SLS009 is in the body long enough to kill cancerous cells, but not long enough to cause severe toxicities as other CDK9 inhibitors have shown. High potency, really great responses that we have seen to date, again, with a favorable safety profile. What's the rationale for AML as the lead indication for SLS009, and can you briefly review the completed phase II trial in relapse refractory AML and the results that were delivered? Yeah. Briefly, since we're focused on AML, it's a very important point here. We're targeting arguably the most urgent and biggest unmet medical need in AML at this time, which is resistance to venetoclax-based regimen. There's virtually no AML patients that I've said before in the U.S. who do not get venetoclax at some point. Unfortunately, there are also no patients, and again, unfortunately, who do not develop resistance to venetoclax at some point. When they become resistant, the median survival is around two and a half to three months, which is well documented. Venetoclax in combination with hypomethylating agents is really the staple of AML treatment across patient categories. SLS009 exhibits really strong synergies with venetoclax and has the ability to both improve response to venetoclax or even convert complete resistance. We have seen that because SLS009 leads to suppression of MCL1 expression and can basically kill those cancerous cells. venetoclax depends on Bcl-2. Some cancer cells, they depend on MCL1 and Bcl-2. We have really a twofold assault, if you will, by both anti-apoptotic agents at the same time and with azacitidine being part of the regimen. We also allow Noxa release enhancement. Thus, we're really increasing pro-apoptotic effect. It's basically a triple hit with a high response rate. That also holds true, by the way, in TP53-mutated patients as well. To your question, what we have seen in the phase II study, quite unprecedented in relapse refractory patients, response rates as high as 50%, median survival of 8.9 months. That also includes in patients with TP53, we surpass the historical benchmark of around two and a half months of survival. During phase II trial and when you expect to release data? The phase II study that's currently ongoing is in two cohorts. Front line, high risk, newly diagnosed AML patients. They're randomized one-to-one to azacitidine only versus azacitidine plus SLS009. The second cohort is the one that is after two cycles of venetoclax. If there's no response or they're refractory, they will also be randomized to or continue to azacitidine, we'll put SLS009 on top of that. It's a study. CR is the primary endpoint. We'll follow for survival, we expect top-line data by year-end. What's the bar to moving SLS009 forward in each of the cohorts, and what would be the next steps following this phase II? Yeah. I think with our extensive work that we have done in the omics work, we want to see high response rates and high CR rates in excess of 30%, 40%. I'm as bullish to say even 50%, hopefully. With the top-line data that we will announce year-end, we're extremely excited really of the potential to use the phase II data towards an accelerated approval path and/or quickly move into registrational study, especially based on the predictive biomarkers we've identified, and also move into solid cancer, which we have also discussed with the agency. Do you have a sense of what the bar would be to support a filing based on the phase II data? Yeah. I think certainly in the patient cohort that's refractory, who by default have a 0% response rate because they are refractory, any CR rate of 30% or above I think could pave the way, and then you would have discussions, obviously, with the FDA if that would suffice or if we have to expand the cohort with additional patients or like. I think our predictive work, Philip, that we have done is really quite exciting how we can really identify the right patients. Great. With that, I think we're out of time. Angelos Stergiou, thanks so much for coming on today. Really exciting times for SELLAS. We look forward to the upcoming pivotal data. Thanks so much, Phil. Thanks for having us.