Good morning, and welcome to the Gritstone bio KOL event. At this time, all attendees are in a listen-only mode. A question and answer session will follow the formal presentations. If you'd like to submit a question, you may do so by using the Q&A text box in the middle of the webcast player. As a reminder, this call is being recorded and a replay will be made available on the Gritstone bio website following the conclusion of the event. I'd now like to turn the call over to Andrew Allen, Chief Executive Officer of Gritstone. Andrew? Thank you, Tara. Good morning, everybody. My name is Andrew Allen. I'm the CEO and one of the co-founders of Gritstone bio, and it's my pleasure to welcome you to our webcast today. Our lead cancer program, GRANITE, is a personalized neoantigen immunotherapy that is currently in a randomized phase II study for metastatic microsatellite stable colorectal cancer. We will have mature PFS data from that trial this quarter. I will not be talking about those data today, however. We today are going to be talking about the disease that we're seeking to address. I'll be making some forward-looking statements through this presentation, and I refer you to our filings with the SEC for further information. Now, I'm very excited today to be joined by two experts in this field. We're holding this webcast because there aren't many people working in frontline metastatic colorectal cancer. There aren't actually that many people working in colorectal cancer at all, but the majority of the trials are in second or often third-line patients, and very few are performed in frontline metastatic disease, which is where our trial is set. So there is probably a relatively low level of awareness of the nature of this disease and of how treatment currently works. And that is the objective today, is to educate our audience around this disease, so that you are ready to receive our data later this quarter. Howard Brown is both a colorectal cancer patient and a highly recognized advocate in the colon cancer arena. Along with having his own book and podcast and being an active speaker in the colorectal cancer world, Howard serves as the chairman and president of ColonTown, an organization that serves over 11,000 patients and caregivers with colon cancer. I'm thrilled that Howard's been able to join us today, and he'll talk about the patient perspective on this disease. After we've heard from Howard, I'll introduce Professor Randy Hecht. Randy is a professor in the UCLA GI Oncology Program and a leading investigator in solid GI tumors. He's also an investigator in our phase II GRANITE study, and you'll be hearing from him shortly as well. Now, before we get to the conversation, a quick bit of background. Gritstone bio was founded in 2015, so 9 years ago, and the fundamental hypothesis was that we would develop potent vaccines to introduce neoantigens into patients with cancer, to drive neoantigen-specific T cell response, to augment that T cell response with checkpoint inhibitor therapy, and that that combination of approaches would or could lead to real clinical benefit for patients with high levels of unmet need, particularly those with so-called cold tumors that do not respond to checkpoint inhibitors alone. That original hypothesis is the one that we have been focused on lo these 9 years, and we're about to test them now, in a randomized phase II prospective trial. Nine years can be looked at as a long time or an incredibly short time, and I'm excited that we're about to learn together whether this approach does indeed add benefit to patients who really do need something better than they receive today. We'll hear more about that in a moment. To develop this, we have produced two core technologies. Firstly, identifying which mutations function as neoantigens, and we do that using an artificial intelligence model. This was hot and new back in 2015. It's obviously now old hat, but nonetheless, has become standard of care and is a very elegant approach to this interesting problem. But of course, you then have to give the neoantigens using highly immunogenic vectors. We've developed so-called heterologous prime boost system, which I'll touch on later in the presentation. The area we ended up focusing on is metastatic microsatellite stable colorectal cancer, and that obviously is why we'll be talking about this today. We'll, Randy will give you some more information, but this is the second leading cause of global cancer deaths. We've done work in this disease already, of course, in a phase I/II environment in patients with advanced disease, and we saw de novo induction of neoantigen-specific CD8 T cells in over 90% of our colorectal cancer patients. We have received FDA Fast Track designation for metastatic microsatellite stable colorectal cancer. The mature PFS data from our ongoing phase II study are expected this quarter, as I said, and overall survival data are expected in mid-2025. We are working with patients and their advocates to shape our program to address this problem accurately, and that means both ColonTown, which Howard represents, and the Colorectal Cancer Alliance in particular. The pivotal study that we aim to start next year, obviously, will follow on from what we hope will be a positive phase II study. And currently, we're finalizing the study design. And our plan, of course, will be to go to the agency as early as the fourth quarter of this year for an end of phase II meeting to review the PFS data and agree on the key design elements for a pivotal study aiming to launch next year. And it is expected that the primary endpoint will be either progression-free survival or overall survival. Now, there are 1.3 million people living in the United States who have had experience with colorectal cancer. Happily, some of those patients, a decent number, are cured, but unfortunately, the majority will succumb to their disease. There are over 150,000 new diagnoses of colorectal cancer in the U.S. each year, and although about a third present with metastatic disease, another large group of about a third will progress from earlier stages to develop ultimately metastatic disease. This is an expensive problem for the United States and for our economic system, but the real expensive cost is in the form of the impact it has on patients and patients' families. Unfortunately, what we've learned over the last few years is that simple checkpoint inhibitor therapy has really very little effect in microsatellite stable colorectal cancer. And the majority, the large majority of patients with colorectal cancer that is metastatic, have this MSS form of disease. And Randy will touch on the successes we've seen in the roughly 5% that have the so-called microsatellite unstable disease or MSI-high disease, and that's, I think, hopefully a beacon for how effective immunotherapy might be if we can harness it effectively. The current data suggests that just giving checkpoint inhibitors alone does not get you there. So with that brief introduction, I'd now like to hand over to Howard. And Howard, it'd be great to hear from you. Of course, today, our audience, I'm sure, is excited to hear your perspectives. So let me hand over to you and hear, first of all, perhaps your your story, please. Sure. Thank you, and good morning, and just an honor to be here and to be able to share patient and caregiver perspective as to what you're doing and what all this means. So first of all, I also, I am just an honored patient champion of the Colorectal Cancer Alliance and chairman of the board of the PALTOWN Development Foundation, which is the fiduciary 501(c)(3) for ColonTown. And we've now, since 2011, aggregated the patient and caregiver journey, and be able to aggregate that and be able to guide them and walk with them upon diagnosis through treatment and through survivorship. I'm gonna step back my story, and what you're looking at is my family, because when you get diagnosed with cancer, it's not you have cancer, it's we have cancer, and we have cancer as a family, as a community. You're looking at milestones that I have been able to achieve that many that who the cancer burden has become too great have not. So I did get to actually see my daughter graduate from high school. I did get to see her graduate from the University of Michigan, and I hope someday and realistically walk her down the aisle. Mm-hmm. This is the real life that cancer has on people. I wanna take you back now to 1989, just for a brief moment. Mm-hmm. I was diagnosed with stage IV non-Hodgkin's lymphoma. There was no term called young onset. I got a book on cancer. The Internet wasn't existing, no cell phones, and computer use was just starting. This was an analog time. I call it my deer in the headlights time. The therapies that they were throwing at me as a, as a young man were... weren't working. I was relapsing, and I got a piece of good news in February of 1990, that my twin sister was an exact HLA match. That's like hitting the lottery. That's a one in 25,000 chance. On May 24th, they brought a purple bag into my isolated room at Dana-Farber Cancer Institute, called an allogeneic bone marrow transplant. That transplant could have killed me. It could have given me extreme side effects, which it did a little bit of graft-versus-host, but it ended up being the trigger to change my immune system and save my life. I immediately went to Arm 1, Patient 4 of a trial called Interleukin-2, the grandfather of immunotherapy. And what I'm told and what worked was that it injected super killer T cells into strengthen my immune system and keep any recurrence from happening. And when I finished that six-month weekly treatments, I was released to surveillance, and I got my life back. Grandfather of immunotherapy, a bone marrow transplant, gave me my life back. I just want that to resonate right now. It doesn't happen that often. So I am blessed and grateful and lucky to have that experience. Now, I got my life back. I got married. I became a big brother. Frozen sperm gave us a daughter. I went to Silicon Valley and took companies public. I went on great vacations, and then I moved to Michigan to bring the family together, my twin sister and my wife's from Michigan. And unfortunately, at age 50, with really no symptoms, I got diagnosed after colonoscopy with stage III B colon cancer, and lightning struck twice. Immediately did surgery of a hemicolectomy or a colon resection of 13 inches of my colon, chemo port installed. And then this is what happens with most colorectal cancer patients, is that we get pushed right to chemotherapy. And these chemotherapies, as Dr. Randy Hecht will tell you, are over 20 years old, and many of them come with significant side effects. They change your life emotionally, physically, financially, and in relationships, and they mostly don't work. And so I was very early on in 2016 and 2017 to get my biomarkers. That means my tumor and my blood were tested, and unfortunately, I got the news that I was microsatellite stable instead of MSI-high. I was ineligible for immunotherapies that were out there and helping people... And so I went through 12 Rock 'Em Sock 'Em cycles of first-line therapy. It didn't work. Cancer had spread outside of my colon. I did another colon resection. I did a failed clinical trial, and then a year later, I went metastatic. Colon cancer spread to my liver, my peritoneum, my omentum, called the stomach lining, and my bowel. Well, you don't want to Dr. Google that because that gives you a 4% chance of living the next six-12 months. And then they start to throw the kitchen sink at you. These are second-line and third-line therapies that have been around for a while, and they come with a whole host of their own side effects. I was very fortunate to get some stability and a little bit of regression, shrinkage. You can ask George Costanza on Seinfeld about that. But that, for a colon cancer or colorectal cancer patient, is great news. Stability and regression is where you want to go; progression is not where you want to go. Well, I will finish up this story by saying that I actually, in ColonTown, became a very smart, not a deer in the headlights, but a smart patient to be able to be collaborative. The missing piece of precision medicine is the patient and their caregiver, and don't forget the caregiver, because they are collecting data and helping with the decision-making through this frontline journey when you're facing death. I ended up finding this crazy surgery called cytoreductive hyperthermic intraperitoneal chemotherapy. You're looking on the slide of 90 staples, 13.5-hour surgery, where they debulked or took out dead and live cancer cells, my gallbladder, shaved my liver, took out my stomach lining, cut my bladder, and I woke up in the ICU and pressed the morphine drip button. It's an intense surgery, and then they poured hot chemotherapy inside my pelvis and abdomen to gain absorption with the hopes of killing micro cell cancer. Well, I'm one of the fortunate ones. It's worked for me. I, in September, will have five years of no evidence of disease, and I've dedicated my life back to educating patients, okay? The statistics don't lie. 153,000 diagnoses, 100,000 people in the U.S., and way more internationally, will be in treatment this year, and 53,000 will die from this disease. We must get more solutions. The last thing I will tell you that solving this jigsaw puzzle is a team sport. Patient advocacy, care partner advocacy with the industry, with the oncologist, we can do this. The last thing I will tell you is that solving this is a team sport, and I am grateful that the progress that Gritstone is making because it gives us all hope, and that's what we need to get one more day, one more week, one more month, and one more year to make memories and live our lives with our families. Thank you. Thank you, Howard. It's such an inspiring story, and unfortunately, of course, you've had to be lucky to be here. I'm obviously thrilled that you have been, but as you said, so many are not. You told me something the other day about when you joined ColonTown at the beginning of your colorectal journey, you had a large group of contemporaries. Could you just remind us of what happened to your friends from those days? So my wife and I joined ColonTown in 2017. We were initially a cohort of 25 new people entered in, I as the patient, and my wife, Lisa, as the caregiver. And some people just have a caregiver, some people, the patients themselves, get involved with ColonTown. And we were learning and getting smart together because of the people that had the frontline journey that were 5 or 10 steps ahead of us. These are the people, the survivors, and I am the only one still alive out of the cohort that joined then. It's just the fact that this disease is a very difficult cold tumor, and we need more effort and more dollars of research and more companies trying to solve it to save lives. Yeah. So you, you received high-dose interleukin -2, which obviously was the, as you said, the grandfather of immunotherapy. And interestingly, that was actually the first program I worked on when I moved to the Bay Area at Chiron, who were the owners of that, asset back in the day. Which obviously received approval in the U.S. for metastatic melanoma and for renal cell carcinoma. And we've learned an enormous amount in the subsequent 20 years, actually now 25 years since the drug was approved, about how it works. High-dose IL-2 was pretty tough. Do you, do you remember much about receiving it back then? I do. I used to go in and I have to tell you, you spend a lot of time with your infusion nurses and trying to find out what's up, but they would change a cartridge, and on a weekly basis, I wore a fanny pack. I will tell you that relatively handled it reasonably well, and the fact is that I kept getting great blood tests, that my immune system and my T cell count was just going off the charts. It was working for me. Yep. I already... You know, when you have to be so resilient as a cancer patient because lots of stuff comes your way, and side effects are no fun, and we have to deal with that, and there's lots of them. Yeah. We've obviously learned a ton, and so much of what we do in biotech obviously is learning from your ancestors, your precedents. You know, there were a couple of giants working on that. Steve Rosenberg, of course, was instrumental. And I'll just namecheck Gwen Fyfe, who many of you, I think, will know, who was the head of development at Chiron and secured approval of that drug, subsequently at Genentech. She went to Genentech and did great things there. We learned about how that drug activated T cells. We knew nothing about what those T cells were recognizing, and that was the key insight, actually, that was delivered back in 2014, that really triggered us to start Gritstone, which was this insight that it was probably mutated neoantigens-... were the key targets for those activated T cells. That's really the catalyst for this particular round of innovation. Now we've learned to use perhaps better immune therapies, and obviously, the PD-1 antibodies have been transformational to solid tumor oncology. One of the reasons they're so attractive, I think, is because of their relatively benign toxicity profile compared to chemo. Could you perhaps speak to that issue around tolerability and how, you know, the impact that chemo has? It's easy to be glib about it, but for patients, obviously, it's a very different story. Perhaps you could just speak to that for a few moments. I would like to. The colorectal cancer patients that, in our community that are, microsatellite, MSI-high, they actually are detailing tolerable side effects. They can go to work, they can play with their children, and they're not the systemic chemos, which are- can... and with steroids or with, you know, giving you, you know, obviously, hair loss, nausea, re- just, s- skin, you know, rash. Really difficult things to navigate through, and we're- and we have some antidotes for them and some, some things, but it's, it's a very different path for microsatellite, stable patient than an MSI patient on immunotherapies. The immunotherapies aren't 100% working, but they are giving people stability, and some people are getting to no evidence of disease. They're just two different paths. One is the difficult path, and the other one is a little more tolerable, of course. Yeah. Yeah, so obviously, high efficacy with very good tolerability is the goal, and that's obviously what we're chasing at Gritstone. Thank you, Howard. I'm sure we'll come back at the end, but let's move on now, if I may- Thank you ... to Randy. So you've heard the patient perspective. It's with great pleasure that Randy Hecht, who is very seasoned and wise, solid tumor GI oncologist at UCLA, known to many. Thank you for joining us today, Randy. Let me move to your slides, and I'd love to hear your perspectives on the clinical unmet need and treatment landscape for this disease. Thank you so much, Andrew. Can you hear me? We can. Okay. Yeah, I run the GI cancer program. I love the word seasoned. It makes it sound though I'm like a steak, but I'm not done yet, and so they are at least well done. It's really hard to follow Howard or, to be honest, any of our patients, and so we see it from the other end. Though, of course, just being an oncologist and being an American means that almost all of our lives have been touched with cancer in one way or the other, and that's one of the reasons we work for the patient. We're trying to get this better because, as Howard was saying, there's a small group, 3%-5% of people, who do extraordinarily well, and the others not as well. While we help them, we have clearly an unmet need. Colorectal cancer is the field where I've been working for the last more than 20 years, originally as a gastroenterologist and then as a medical oncologist. It's the second most common cause of cancer death, number 3 in men, number 3 in women. It's also the second most common cause of cancer death worldwide. Next slide, please. However, I am so glad that Howard has beaten the odds, but he has beaten the odds, and we know that people with early cancers, which also explains why, you know, it's important for early detection, patients do well. Patients where it's spread to local lymph nodes, there's still a lot of those people, unfortunately, will end up dying of their disease. People who have metastatic disease, either upfront or, as Howard was mentioning, if it becomes metastatic later, other than a relatively small number of people who can have liver resection, even a smaller number of people who have cytoreductive surgery and heated intraperitoneal chemotherapy, almost all of those people will succumb from their disease in the first, in the first five years. Next slide, please. One of the most distressing things that I see, you know, when I go to my clinic every several times a week, is that for reasons that we don't understand, is that while we've been doing a better job in screening and actually rates of, of cancer death have fallen some in older people, I see patients in their 20s and 30s and 40s all the time here at UCLA. In fact, for reasons that we don't understand, colorectal cancer is significantly increasing in young adults, which is a group that isn't really thinking about it. It's a group that we don't screen. It's a group where symptoms are ignored, and by the end of this decade, colorectal cancer will actually be the most common cause of cancer death in patients, in people under 50 in the United States. This is... All colon cancer is terrible. Telling someone who's in their 20s that they have colon cancer is even worse. Next slide, please. Howard, with a broad brush, was sort of painting what we do with colorectal cancer, and then I think one thing that either he or Andrew said is, I have been fortunate enough to be around, through my career as we've developed really the modern treatment of colon cancer. The problem is that most of the benefits, most of the advances, I'm sorry, that we've had, some of which, including bevacizumab, where the first trials were done at UCLA, anti-EGFR, I ran a lot of the early, actually most of the U.S., panitumumab trials is that those were done in the 2000 s. Other than that small group of patients that Howard was mentioning, 3%, 5%, who have a really different disease, a different way to colorectal cancer, who really respond extraordinarily well to checkpoint inhibitors, that things have not changed that much. We have some newer drugs, but I would say that they're modestly active. First line, so what that means is people who are newly diagnosed with metastatic disease, whether it shows up metastatic or whether it comes back as metastatic, are treated with what's called a fluoropyrimidine. The classic one is 5-FU, which is older than I am, and it's still an active drug. We've learned better ways to give it, but that also tells you the need for new, you know, the need really for new approaches and new drugs. It's usually combined with either another drug, oxaliplatin, or irinotecan, which once again I worked with in the late 90s before it was approved. And that has not changed that much. We can use an antibody, either an anti-VEGF antibody like bevacizumab, also called Avastin, or an anti-EGFR, panitumumab or cetuximab, which would be Vectibix or Erbitux, but that has not really changed. And though I think that median overall survival has increased a little bit, a lot of that has to do, actually back to it, back to Howard, with better supportive care rather than you know really changing the drugs around. In second line, it's a little bit depressing, which is that what you didn't get first line, you ended up getting second line. I run large randomized trials in first line, but also in second line, and people in particular may have a lot of tumor shrinkage and benefit in first line. We do much worse in second-line therapy, so that's patients who have already progressed on first-line therapy. We are beginning to learn about targeted therapies. One of the great advances, if you can say that, has been knowing which patients not to treat with certain drugs, like that patients who have KRAS mutations in their tumor should not get anti-EGFR therapy. We are a little bit late and also not as successful to truly targeted therapies as our colleagues in melanoma and particularly lung cancer. BRAF mutations are found in roughly 5%-10% of patients, depending on the stage. With colorectal cancer, they tend to do really poorly. We have statistically significant improvements in outcome. They still do relatively poorly. In the salvage setting, we actually have a number of modestly active drugs, and it hurts me to say that there are occasional people, there are always people who do well, but overall, these drugs are relatively modestly effective, whether they're targeted therapies like regorafenib, fruquintinib, which also hit VEGF, whether they're sort of more unusual cytotoxic chemotherapies, such as TAS-102, also called Lonserf. But basically, you know, none of these actually lead to the long-term benefit. Because you ask what you end up getting, an oral session at one of our large meetings is a small improvement in outcomes. What we're really looking for are these great improvements in outcome that Howard and the people on ColonTown really, really want. Next slide, please. So this is actually just going back a little bit, which I think Andrew and Howard and I have talked about, which is really that these curves, which are clearly, you know, kind of smoothed. But the problem is that even patients who do get initial benefit with chemotherapy, and that's actually what cytotoxic chemotherapy does. The majority of people get some benefit up front. The problem is that it's not particularly durable, and those are the dark lines that you can see. And whether or not it's chemotherapy with fluoropyrimidines like 5-FU or capecitabine or multidrug chemotherapy, the problem is that it wears out. On the other hand, in the 3%-5%, and this is what is so promising for us in trying to get immunotherapy to work in the 97% of people who are, microsatellite stable or don't have a lot, as many mutations, is that checkpoint inhibitors such as nivolumab and ipilimumab, really, you know, we're not quite at saying cure, but these people have long, long-term benefit, and that's what the patient wants. Next slide, please. So as Andrew alluded to upfront, well, why don't we just use, why don't we just use these checkpoint inhibitors, immunotherapy in colorectal cancer? And the problem is, is that they don't, they don't work. I remember when my colleague, Eddie Garon, whose office is next door to mine, presented the KEYNOTE-001 trial in lung cancer, and they do work in a large percentage of people with lung cancer. But when you take these same drugs and you add them to, you know, either as maintenance or, or together in other, you know, in other settings, in patients with colorectal cancer. And maintenance, I'm just gonna talk about just for a second, but when you add that, there really doesn't seem to be benefit. So clearly, colorectal cancer is different than lung cancer. When we use maintenance, it's back to what Howard was saying.... People can tolerate, generally, chemotherapy for a while, and it shrinks tumors, and it makes people sometimes feel better, actually, if the tumor is causing problems. But in particular, first-line chemotherapy is hard to continue for a long time. So what happens is that patients end up getting at least stabilization or improvement in their disease, and then they go into what's called maintenance, which is kind of a lower impact, usually with a drug such as capecitabine, which is turned into 5-FU. And so that's a really interesting place to look at drugs because patients are generally doing well, and we know that patients who are doing well are more likely to benefit from immunotherapy, but we know that we're kind of waiting for the other shoe to drop. Next slide, please. I'm a GI oncologist rather than a rather than an immunologist, and so, like, but I'm very fortunate to work with really good immunologists, both at UCLA and also at Gritstone, who have helped me understand this as we've tried to find any way to get to the immunotherapy revolution. My colleague, Tony Ribas, who's at UCLA, who's a well-known melanoma researcher, kept telling us that the T cell is the drug. And I think, Andrew, you kind of mentioned that before, whether it's stimulating it through IL-2, like how it had, or whether it's through vaccination, or even cell therapies, which have not been successful yet in really in colorectal cancer. The question is: How can we make the T cell the drug? And this cartoon looks at various sort of possible tumor situations. Now, there are some tumor cells, particularly pediatric, tumors, that have very few driver mutations, very few new antigens. Those are basically altered proteins or other chemicals that theoretically the immune system could identify, but is not. Otherwise, the patient wouldn't show up in my office. But if they don't have those, then these types of approaches are really not gonna work. Now, there are also tumors that have neoantigens, and the problem is that we're not really getting the T cells to recognize them, and this is one place, such as actually what I work with, which is MSI-low, colorectal cancer, but there are other major malignancies. Pancreas cancer is also the third most common cause of cancer death, with even a more grim prognosis. If you do have neoantigens, neoantigen-specific immunotherapy has the chance to get de novo or new T cells that recognize those neoantigens that are just not being recognized now. Even more exciting is that there are some tumors that... I won't say more exciting, but also exciting, is that there are also tumors where there is a T cell response. It's just not doing what it's supposed to do. Those tumors often get some benefit from adding checkpoint inhibitors. But obviously, if you can sort of amplify that, or if you can add and amplify that by adding de novo T cells, once again, bringing more soldiers to the fight, those patients who, particularly in upper GI cancers that I also take care of, they may get some benefit, but they don't get the long-term benefit either. The people who are already at the table for the immunotherapy revolution, such as melanoma and MSI-high, already have this robust immune response that we hope that neoantigen-specific immunotherapy can replicate. Next slide, please. I get the privilege of talking about the approach, and it's coming in as a, once again, as a GI oncologist who, like all of us, have had to learn to be immunologists. Is that the approach that we're doing actually, I think is quite clever. So when you look at checkpoint inhibitors, as Howard mentioned, PD-1 or PD-L1 are well tolerated. Just by themselves, they have not worked. They're probably necessary, but not sufficient. They also, as Howard was talking about, that they have rare, severe, but generally are extremely well tolerated, which allows you to combine with other things. On the other hand, anti-CTLA-4 checkpoint inhibitors, such as ipilimumab, particularly when given systemically, have a lot of toxicity. On the other hand, probably by priming the immune response in local lymph nodes, the question is: Do you need to get this, you know, systemic, these systemic side effects in order to get really an immune response that you might need locally? And being able to give, as in the GRANITE trial, being able to give local CTLA-4 has a lot less toxicity and actually may be, you know, putting it right where the neoantigens are, right where you want it. And this is one we have found that you don't get the systemic side effects by doing it this way, and there's the promise of having even better, and even a better immune response. Sorry, it's early here on the West Coast. Next slide, please. So to summarize my part, which is that colorectal cancer, which those are the patients that I see every day, and it is still the second most common cause of cancer, of cancer-related death, not just in the United States, but worldwide. We know that while after this burst of activity with targeted therapies, better chemotherapy, really, things have stagnated for our patients since the mid-2000s. Think about how things were in the mid-2000s with your computer. This is how it is with, unfortunately, with my patients. We certainly help patients, but when, when patients say what they want, what they want is not help for today or, or improvements measured in months or hazard ratios that are minimally better. They want long-term benefit. As I said, we're helping about 5% of patients, but the unmet need-... is in the 95%, and I'm very excited about this vaccine approach. It makes sense to be able to get that immune response. So next slide, please. I think I turn this over to you, Andrew. Thank you. You do. Thank you so much, Randy. I appreciate it. And again, we'll be opening to Q&A shortly for questions on your aspects here, but thank you for that. So just to remind you, the notion here is that there are patients out there who have tumors that contain mutations, not huge numbers of mutations, and therefore they have neoantigens, but not huge numbers of neoantigens. There appears to be a somewhat proportional relationship between the probability of developing a neoantigen-specific T cell response naturally, and the number of mutations in your tumor. And this obviously was described by Naiyer Rizvi and his colleagues in a Science paper back in 2015, showing that the probability of responding to a checkpoint inhibitor related pretty, directly proportionally to the number of mutations in the tumor, hence the development of tumor mutational burden, or TMB, as a predictor of benefit from checkpoint therapy. So there are patients who have mutations, they have neoantigens, but they're not lucky enough to have developed a natural neoantigen-specific T cell response. And when you treat those patients with checkpoints, nothing much happens. If you're lucky, toxicity, if you're unlucky, and this is probably microsatellite stable colorectal cancer. So our thesis was, let's give them neoantigen-reactive T cells by sequencing the tumor, by plugging all the mutations into our artificial intelligence prediction model called EDGE, and then figure out which mutations will function as neoantigens. And then we'll put 20 candidate neoantigens into two different vaccines, an adenovirus priming vaccine, and a self-amplifying mRNA boosting vaccine, and we'll give those to patients together with checkpoint inhibitor therapy. And we will then prime de novo neoantigen-reactive CD8 T cells, probably CD4 T cells as well. These will infiltrate tumors, kill tumor cells, and drive clinical benefit. That's the hypothesis that we've been testing these last several years in the clinic. And we've been able to check the boxes or join the dots for all of those different scientific steps that we think are necessary for clinical benefit. So first of all, are we able to accurately identify which mutations create neoantigens? Simple answer is yes. We published that in Nature Biotechnology back in 2018, validating the approach using some cell therapy data from Steve Rosenberg, which is obviously a pretty stringent test because it's totally orthogonal, and yet we were able to predict in silico, which mutations would act as targets in his system of TIL-based therapy. So we can check the box here. We're able to predict which mutations act as neoantigens. When we then vaccinate patients, are we able to generate a de novo T cell response? And we've shown that the answer to that is yes, and we published a fairly extensive, translationally rich paper in Nature Medicine in 2022, showing all of these different elements. We showed that the tumors at the start were cold. There was no evidence in blood of neoantigen-reactive T cells in most of our patients. After vaccination, however, we could see the induction of neoantigen-reactive T cells in the blood. We could see those T cells trafficking to tumor and expanding. And critically, we could see that those T cells could lyse tumor cells. We saw that with drops in biomarkers like CEA, the standard protein biomarker for some patients with colorectal cancer, and then excitingly, we saw it with circulating tumor DNA, a newer biomarker that's perhaps relevant to more subjects. And we saw from single-arm studies, an association between response measured by ctDNA drop or CEA drop, and apparently extended progression-free survival and overall survival. And of course, in a single-arm study, those are associations. You can't establish causality, and that's why you run randomized studies, which is what we've been doing for the last three years. And I will just finish. We turned these cold tumors hot. This is all captured in that Nature Medicine manuscript. So the prospective test to see whether, in fact, we really do drive benefit with our therapy, requires a randomized trial with traditional clinical endpoints such as PFS and OS. That's the study that we've been running. So this, many of you are familiar with this. This is our study ten, GO -10, as we call it. Inclusion is untreated, metastatic colorectal cancer, microsatellite stable, and patients who are being considered for therapy with standard of care chemotherapy, either FOLFOX/Bev or FOLFOXIRI plus Bev. A one-to-one randomization with two stratification factors. Everybody then receives the same induction chemotherapy, and upon completion of induction chemo, study treatment is initiated. In the control arm, that is maintenance chemotherapy with 5-FU plus bevacizumab. In the active arm, the test arm, that is the same maintenance chemotherapy, plus the addition of the neoantigen vaccine, plus systemic atezolizumab, and two subcutaneous doses of low-dose ipilimumab. As we've discussed, Atezo and Ipi have been studied in different trials, but extensively in MSS colorectal cancer, with no evidence of meaningful benefit in that population. The primary endpoint for the study, we said, was ctDNA response. The first secondary endpoint was progression-free survival, and then, of course, overall survival. If we look at the PFS data, which of course, is likely to be, considered as a phase III endpoint, either that or OS, depending upon the FDA perspective. PFS, we looked at earlier this year, and the data were immature. You can see a lot of censoring of patients on these curves, but nonetheless, you can see that notwithstanding that censoring, there appears to be separation of curves in favor of vaccine that begins at around six months. And that's what you'd expect, because most patients received five months of induction chemo and began their immunotherapy at around that five-month mark, and it takes, on average, around a month for a primed immune response to form, and therefore, clinical benefit you wouldn't anticipate much before six months. And indeed, of course, looking at these curves, that's exactly what you can see. And the hazard ratio here on these immature curves was calculated at 0.82. Now, the data were immature. There's just no way around that, but there is an approach that one can take to try and glimpse into the future. The way to do that is to split the population into fast progressors and slow progressors. What's emerged over the last few years is that the best predictor of whether you'll be a fast progressor or not is your baseline circulating tumor DNA, titer or level. The simple truism with ctDNA that we're seeing reflected across multiple datasets, multiple tumor types, multiple forms of therapy, is that high ctDNA at baseline is bad, low ctDNA is better, and upon therapy, you want to see ctDNA stable or going down, whereas relentless climb is bad. So taking that insight, we looked at ctDNA baseline in our patients, or we didn't have data on all of them at this time point, but everyone where we had data, and we split around the median, based on the control arm. So we took all the controls that we had. I calculated the median ctDNA, which turned out to be 2%. We split the populations, both controls and the vaccine recipients, into greater than 2% or less than 2% of baseline. And here are the data. So this doesn't look at therapy. This is patients being aggregated independent of therapy. The black line is everybody whose baseline ctDNA was less than 2%. The green is people who are over 2%. And what's striking is the dramatic separation of these curves with a single variable. That's what's so remarkable here, is that just that one test can split patients, not quite perfectly, but it's not far off, into fast progressors, green, slow progressors, black. So now we could ask the question, let's take a look at the treatment effect in the fast progressors. There's no point in looking at the black population here, the slow progressors, because there are no events, so you can't learn anything yet. Unfortunately, they will progress, but at this time point, very few had. The green population, the fast progressors, obviously, many of them have progressed. We can now ask the question, is there an apparent benefit from immunotherapy with GRANITE? And the answer, gratifyingly, was yes, with an apparent strong treatment effect, again, starting at around 6 months, the hazard ratio now calculated at 0.52, medians of 12 versus seven months. Again, these are still early data, so let's not anchor too much on these single numbers. The point is, directionally, we're seeing an apparent benefit from immunotherapy, even in fast progressors. If you'd asked people at the beginning, "Do you think the product will work better in fast or slow progressors?" I suspect most people would say, "We think it'll work better in slow progressors, because they have more time to benefit from the immune response, which takes time to form." Unlike fast progressors, who often are galloping with their disease, unfortunately. So these were very encouraging data, and of course, we're now on the cusp of mature data for the all-comer population. And again, I want to just dispel the notion that I hear sometimes. We are not saying that we're only focused on fast progressors. These data were informative in fast progressors and uninformative in slow progressors. We anticipate having data in everybody, fast and slow progressors, at this next data cut this quarter, and all of these patients need help. Being a slow progressor is better than being a fast progressor if you are a patient, but it's still not good enough, because you almost certainly will progress. Importantly, and we touched on this earlier, with Howard's comments, the tolerability of immune therapy is good. No patients have discontinued study treatment due to a treatment-related adverse event, and the common adverse events are the mild systemic effects associated with any potent vaccine. Of course, given that many of us have had COVID shots, we all know exactly what this feels like. This is 24-36 hours of flu-like illness of varying severity, plus some local arm pain. But it all wears off very quickly, and compared to chemotherapy, of course, this is nothing. So we, in summary, clearly need a shift in the treatment landscape for colorectal cancer, even as early as frontline colorectal cancer, because the current treatments simply are not good enough for the many patients who succumb to this disease each year. It has been a difficult space to work in because we've seen so little success. For the reasons I've described, we saw signals in these patients, third-line colorectal patients, in our phase I/II program. We've seen signals in newly diagnosed metastatic patients we shared earlier this year, and of course, we're excited to see the mature PFS data imminently this quarter, together with the mature survival data, which we anticipate having in the middle of next year. If we're successful in metastatic colorectal cancer, a tough cold tumor, then obviously it opens the door to many, many other options. At this stage, we would be resource limited and of course, hope to have a partner to help us access what potentially will be these, large number of opportunities in many other tumor types, hot and cold, both in the newly diagnosed metastatic setting. But of course, there's a lot of excitement about treating earlier in the adjuvant or maybe even the neoadjuvant setting. We can perhaps talk about that in Q&A. With that, I'd like to turn the call over to the operator, for our Q&A section. Thank you. Great. Thank you, Andrew. So at this time, we'll be conducting a question and answer session with our speakers. For those on the webcast, if you'd like to submit a question, please use the Q&A text box at the bottom of the webcast player. And to our analysts who are joining us live, please raise your hand to indicate you have a question. So please hold for a brief moment while we pull for questions. So our first question comes from Marc Frahm at TD Cowen. Please go ahead, Marc. Hi. Thanks for taking my questions, and thanks, everyone for putting on the presentations. Those are great. Maybe one from an epidemiology perspective, Dr. Hecht, I mean, you mentioned this kind of overall trend of rising incidence among younger patients of CRC. Is that coming with any sort of changes in some of the demographics of mutations and, you know, and other things that might impact outcomes here in subsets of, important subsets of, colorectal cancer, and kind of what impact that may have on the longer-term direction? And then I have another question after that. Sure. Do you want me to go ahead now and then... So that's actually a real, you know, sort of active field of research. So, you know, back to the whole MSI, MSS, it used to be that the young patients that we saw were actually mostly, you know, patients who had inherited predispositions, such as what's called Lynch syndrome, and those patients really are MSI high. They have the thousands of mutations. What's really striking from epidemiologic standpoint is that, one, that the this tsunami of young people that we're seeing, one, they do not, they're not MSI, they're MSS. Second of all, it tends to be distal, so they're usually sigmoid or rectum. And there's ongoing, you know, there's ongoing research as to whether or not there really are, I mean, it depends on the data set. What I would say is that there may be some differences, but I think that from an approach standpoint, these still are patients who, you know, who are MSS and who could, you know, benefit by something that would increase, sort of recognition of neoantigens. They don't seem to be that different than older people. That's a separate discussion, is why epidemiologically, and probably not germane to this, but epidemiologically, why suddenly, not just in the U.S., but, you know, people in Western societies, why young people are beginning to get, distal, rectosigmoid cancer. But it's not like all totally different, and it is not inherited, and it's not MSI high. These patients, from an immunologic standpoint, are MSS colorectal cancer, just like others. Did that help? Yeah, very helpful. Thanks. Then, maybe, probably, again, your thoughts on this from a kind of clinical, one, your experience with patients, but also just clinical execution perspective, and then, look at Howard's perspective, too, as a patient. Just, you know, obviously, the side effects here are, you know, relatively mild or relative to some of the other options, but there are kind of other impacts on quality of life and burden, you know, kind of logistical issues around the care. Just, how do you, given those, like, how do you view what's a meaningful change, initially in PFS, but, you know, and then ultimately, as we get the OS, like, what's the kind of minimum that you would want to see to kind of justify, you know, going down this path? Is it really about moving those medians, or are you much more interested in whether there's a tail effect or not? That is a great question, and I would say maybe all of the above. And so that it shows that you've actually been thinking about this. So on, you know, on one hand, there's an old colleague of mine, Lee Ellis, who's at least for now, I think, still at MD Anderson, and who said, "Well, if we're going to have a first-line trial, you know, the minimum, you know, difference in, particularly in OS, should be a hazard ratio of, you know, of 0.8. And anything, you can overpower a trial, but anything less than that probably isn't that helpful." But I think as you pointed out, you know, there's a big difference between a curve and 1,000 patients and what the patient wants. So on one hand, yes, we want people to do better, and if we can improve the PFS, you know, it helps people. Don't get me wrong, better PFS means more time with your family. Better PFS means more time without the cancer causing problems. So there's no problem with looking at median PFS. The flip side, and this has always been, if you want to know what the patient who's sitting in front of me wants, they don't want an improvement in PFS. They want to be on the tail of the curve. They want to be on the right-hand side. They want to be like Howard. They want to be Howard. And so they want to be that person. So I think both of those-... I think what you brought up is that, you know, I'm a clinical trialist, and we look at these things, and we look at medians, and we look at hazard ratios. But really, the patient says, "I don't want to be a hazard ratio. I want to be in that tail of the curve." Hopefully, with immunotherapy, you can do both. If your tail is big enough, tail of the curve is big enough, then you start moving the PFS medians, and you move OS medians. Both of those are good, and it helps my patient, but I think, but for the first time, we're actually talking about the chance to get long-term survivors. Was that helpful? Yes. Yeah. Thank you, Randy. Thanks, Mark. One last thing is that sometimes these also happen to be what the patient wants and what the agency wants is not always exactly the same thing and what the statisticians want. But you have to satisfy them. If you have an active drug, you'll satisfy everyone. From the patient side, I want to just include that, first of all, we have a really tough side of getting quality of life, okay, versus being in treatment with side effects, and we have such limited options in the later stages of colorectal cancer. We end up repurposing and re-challenging them after a break, and it's, we're just left with no decisions. We've got to hunt for clinical trials, and so quality of life matters, treatment matters, but living matters when you're on the front lines of this. So, this option, as I concluded my presentation, it provides us hope. Yeah, thank you. And I'll just remind our audience that our trial is pretty inclusive. We take, you know, pretty much everybody in the front line, you know, with standard clinical inclusion/exclusion criteria. But, you know, for example, 75% of patients presenting with metastatic colorectal cancer have liver metastases. Those patients are included in our trial, and indeed, the fast progressors pretty much all have liver metastases, and yet we saw treatment effect, apparently, that, that I shared with you a few moments ago. So being inclusive at the beginning, I think is a, is a real value add here, not selecting patients assiduously. Great. One thing that I... Go ahead. No, go ahead, Randy. I was just going to say, one thing that I, I do also like about this is that, as someone who's been doing drug development for a long time, often, what we do is that we take patients who've, as Howard was saying, who've had multiple therapies, who've had multiple modestly active therapies that may have toxicity and also have had their colon cancer, their colorectal cancer continuing to grow. One nice thing about, this approach, and it's, it's, you know, others have also realized this as well, is that bringing this early in, I think is much more likely to get a good immune response and to improve that quality of life. And even with chemotherapy, our biggest benefit is early on. And I think from an... I'm not, as I said, I'm not an immunologist, but we know that from an immunologic standpoint, also being able to give something, and maintenance, I think, is just a great place to test drugs, you know, from all sorts of different reasons, is a great place to be. Patients are not getting a lot of toxicity from multiple other drugs. They've dropped the most toxic drugs. Also, immunologically, this may be better. The earlier, the better, as the immune system is not beaten up necessarily by, you know, by what we do. And so I think that's also sort of an important, you know, something that's really interesting about this as well. Go ahead. I'm sorry. Thank you. We should, we should move to our next question, Tara. Yes. So our next question comes from Mayank Mamtani at B. Riley. Please go ahead, Mayank. Good morning. Thanks for this session. Obviously, patient and clinician perspective, very helpful. Maybe just first for Dr. Hecht, you know, picking on the prior to topics, you know, that were discussed, you know, if you could share the regulatory perspective on, right, 'cause the obviously, for them, increasingly, survival and these quality of life metrics are becoming increasingly important. But was curious if any proof of biology markers or kinetics of separation versus control in this study, you know, is informative of activity. And essentially what I think we're all thinking about is how learnings here could benefit the phase III study design for next year. And then I have a follow-up question for Andrew. Some of these may also be for Andrew as well. So I think regulatory is important. At the end of the day, that if your drug isn't approved, then it doesn't help patients. And partly as a clinical trialist, I work with people, hopefully, that you have to make certain that drugs that have activity, that you don't mess this up and, you know, throw the baby out with the bathwater. What I would say is that on the one thing that from a regulatory standpoint that has been really useful has been that mirroring actually kind of what I was talking about before, from a drug development standpoint, we used to have to say on our patients, that the patient has exhausted all options or has refused. And so once again, you get beaten up. New guidance from the agency allows us really to bring this, you know, early drugs, such as this, particularly ones that, that seem to not have as much in the way of toxicity much earlier in. I don't know if that was exactly your question. As to trying to build the modeling, I'm not a statistician, but the reason you do randomized phase IIs is to actually get your assumptions for your phase III. ... Yes, let me pick up on that. So we obviously are measuring T cell responses in a variety of ways in this trial, and we will explore whether those are related to outcomes. We anticipate PFS this quarter. There's evidence with immune therapy that PFS may underappreciate OS benefits. We'll see the OS data next year. Definitely. Mm-hmm. And then, obviously, the correlative work can follow. It's obviously important that we're now pivoting from- Yeah Clinical biomarkers that you use in early-stage trials, such as ctDNA, to registrational endpoints, such as PFS and OS. Obviously, that's the normal evolution of drug development as you move from phase I/II to phase II/III. So we're on that journey. A PD-1 antibody's been approved in MSI colorectal based on PFS endpoint, so there is precedent, but the FDA will have that determinative view, which we hope to hear later this year. Got it. That's helpful. And just quickly, Andrew, if you could remind us the expectation at this next data cut of fast versus slow progressors. How much... Do you anticipate all of them sort of progressing around the time when you present data next? You know, would you sort of segment your analysis in a similar way? And also, on the OS curves next year, is—like, should we expect you to kind of start, you know, present data in a similar way for OS also? And lastly, is the GRANITE data going to be presented at a major medical conference in fall, winter? If you could clarify that, too. And thanks again for taking our questions. Sure. So we expect to show the mature PFS data in the overall population, because that's obviously how the study was designed. The notion of fast progressors was useful when you're looking at immature data, or potentially useful, but that's not where we're heading with the program. We don't anticipate needing to do anything complicated in terms of patient selection. As I said, unfortunately, this is not progressors, non-progressors. Most people progress. Fast and slow, they all need help. So, the clinical opportunity here, and the thing that patients need, is that everybody needs something better. So we'll be focused on all comers. Now, obviously, there'll be interest in the subsets because we've shared them previously, but I don't think those are going to be particularly important to the future program that we'll be running. OS data, similarly, overall survival in everybody is what you will care about. In terms of medical meeting, that's obviously to be determined. Once we have more clarity, we'll bring that obviously to our audience. Right now, we don't have a clear answer to that question. Got it. Thanks again, Andrew. Yep. Thanks for the questions, Mayank. So our next question comes from Christian Kubasik at BTIG. Please go ahead, Christian. Hi, this is Christian from BTIG. I'm on for Kaveri Pohlman today. Thanks for today's presentation. It's been very helpful. I understand that microsatellite stable patients do not respond to checkpoint inhibitors, and the whole design of this personalized vaccine is to turn a cold tumor hot. With that said, have there been any positive signals from the phase II trial that demonstrate that adding atezolizumab or ipilimumab are contributing towards efficacy in these patients? You know. Yeah, it's a good question. We spend a lot of time thinking about this. The purist obviously would like to run a three-arm study. However, that's time and money, and the study has been run at some level by Roche. The MODUL study that Randy showed a clip from earlier was the big test in several hundred patients of whether maintenance atezolizumab has benefit in this disease. It was highly analogous to our study design, and so it's a very good external control, effectively. And it showed convincingly there was no impact of atezo upon OS or PFS in any patients in that study, and that was a proper MSS trial. There have been studies done using things like durva/tremi, so that's the AstraZeneca PD-L1 CTLA-4 antibody combination, in patients with metastatic colorectal cancer. That was in a later line setting. Overall, there was no significant effect. There was some separation of curves, but on post-hoc analysis, it looked as though that was all actually due to undiagnosed MSI disease, because the patients who benefited had TMBs in the mid-twenties or higher, which essentially is not MSS disease. And at the time that study was being run, we didn't have a standard approach to assessing for MSI. That was not part of the clinical algorithm. So, if you strip out those patients who are likely MSI high, you are left again with no effect, and that's with the combination of PD-L1 and CTLA-4, and that was systemic CTLA-4. So the purist, as I say, would like us to do that trial, but time and money constraints mean that we don't want to do that trial. Frankly, physicians also told us, "We think it's unethical to repeat studies that we've already done. When we know that there's no benefit, but there is always toxicity, we don't think it's an ethical proposition. Got it. That's very helpful. And if I could just add one more, in general, how are you viewing ctDNA as a biomarker going forward? Yeah, great question. So what we saw in our study is that over time, if you looked at ctDNA from the very first blood draw to the very last blood draw, and this is as of, you know, March, there was clearly more patients on the vaccine arm who had ctDNA flat or down over time than on the control arm. So that's the trend you would hope to see, and it's what we saw.... When we focused in on the short-term period just after beginning maintenance therapy, we saw response- so-called responses, apparent responses, per our definition in the control arm, which we didn't anticipate, so there seemed to be some delayed effect of chemotherapy. So that tells us that using ctDNA as a biomarker in this setting has some nuance to it, nuance that we're just starting to figure out. And so, I believe that ctDNA has enormous value as a surrogate, efficacy marker and will be incredibly useful in early-stage oncology trials. It is not currently a regulatory endpoint, and therefore, like many surrogate or biomarkers, we'll be using it in early stages trials, where single-arm PFS data are often impossible, essentially, to interpret, but drops in biomarkers can be useful. We'll be using ctDNA, perhaps in other trials, in other settings, for early trials. But as you move into, later-stage randomized trials, then, of course, you're gonna be using hard clinical endpoints. The one that matters is OS. There's no getting away from it. That is the endpoint that everyone should care about, and the FDA certainly does. PFS may be a valid alternative if it's been shown to correlate with OS, and obviously, that's a judgment that the FDA makes on a case-by-case basis. Got it. Thank you so much for taking our questions. Thanks, Christian. Our next question comes from Catherine Novack at Jones Trading. Please go ahead, Catherine. Catherine, you might be on mute. We're not hearing you, Catherine. So we'll put Catherine back in the queue for now while we figure out her audio issues. So we'll take the next question from John Miller at Evercore. Please go ahead, John. Hi, guys. Thanks so much for taking the question. Great presentation. I'd love to ask about the bar for success in those different patient subgroups that you've identified. So, obviously, the bulk of the signal at the first interim is being driven by those fast progressors who are high ctDNA at baseline. But Andrew, you mentioned you expect to see benefit in those low ctDNA baseline patients as well. Is there a different bar for success in terms of a hazard ratio, in terms of PFS benefit, that we should be expecting from high ctDNA at baseline, low ctDNA baseline, rapid progressors, slow progressors, that will drive your decisions going into larger studies? Yeah, it's a nuanced question, John. Thank you. I think the simple answer is that, again, depressingly, everybody needs help. And hazard ratio, obviously, is a way of moving away from the absolute, which, of course, will vary by risk group, to a relative, benefit. I think you heard Randy speak to that earlier, so maybe let's have Randy speak to his perspective here. I think, you know, an improvement in PFS and OS, you know, whether it's dramatic or modest, I think all will be welcome. But Randy, your perspective. What I would say is that, it's I think we had discussed that this, this sort of ad hoc, fast, you know, fast progressor versus, slow progressor is useful in reading the tea leaves. And I think it's useful in, maybe even trying to get your, you know, what are your assumptions for designing your larger trial? But, you know, I have to, you know, repeat what Andrew said. You know, at the end of the day, even people who do well, eventually progress, and if you wait long enough, they will progress. Now, are there biological differences between those? That's one of the things that, unfortunately, it's kind of an iterative thing, that you need a large trial to go back and look at those. I would not, you know, change my trial, you know, necessarily depending on that. One problem is that you can't predict who's gonna progress rapidly before you go ahead and you do the, you know, do the treatment. So I think this is useful, sort of in your assumptions. The FDA is not going to look at this. The flip- you know, look, we do learn in our markers, and one problem with ctDNA is also there is no standard. You know, there's a standard way to measure, you know, with RECIST PFS. OS, as was pointed out, is a really hard endpoint. There are many, many different ways to measure ctDNA, and that's one reason why we haven't put it necessarily in the guidelines, which is because there are many different ways. So I think it's one of those things, it's ad hoc, it's interesting. It may help you actually get to biology, 'cause clearly there are biological differences. But at the end of the day, you know, you have to sort of help everyone, and so especially you can't, you can't, because you can't predict. I mean, we, we do, you know, have sort of weird biological subgroups in colon cancer that, you know, as long as they're validated. For example, we talk about sidedness, in what I do. Sided, you know, we treat people differently, whether the tumor is on the right side or the, you know, on the left side. As Andrew had pointed out earlier, there are trials, including one, that I'm running, where, you know, whether or not someone has liver metastases or not. But I would just say that this is the ctDNA right now is a non-validated tool. I think it's useful for trying to understand a little bit what's going on. I think it's useful to sort of be the microwave to get some data, you know, and get some results earlier. But at the end of the day, you're gonna have to help everyone. ... Howard, anything to add from a patient's point of view here? Yes. So, it's patient education, and we're in the education business, all of us. And so, we at ColonTown and the other advocacy organizations now have learned to be able to get our tumor burden and get our biomarkers much earlier. Before, it was when you went metastatic, and that is actually too late. So, upon diagnosis, we actually print out from our ColonTown University education site to take into your first and second meetings within the clinic, because you only have 15 minutes with your oncologist, and to ask them immediately to, within, you know, within a few months, to be able to start to collect that in very important data. And also part of screening now, ctDNA is a big part of the screening that will lead you eventually, whether how you get it from blood or from a poop test, is to getting a colonoscopy. And not getting this disease really helps a lot, too, and we're getting a little better on that as well. So that's my opinion, is that it's the patient education, getting that information so that folks like Gritstone can do their jobs. Thank you. Yeah, so I think to condense our response, John, you know, we're lumpers, not splitters, at this point. Once we've got mature data, we can look post-hoc and see whether there's any strong rationale for starting to split out people. But right now, we have to be lumpers, I think. Well, thanks. That makes sense. Thank you. Tara? So I'm going to read Catherine Novack from Jones Trading questions out loud. She's unable to unmute, but her first question is for Dr. Hecht. So you mentioned that you expect PFS may underestimate OS benefit. What, if any, concern do you have with pseudo progression when evaluating PFS with GRANITE? That's actually a good question. So I think there are two parts to that. So one is that in places where immunotherapy works really well, in the long term, sometimes you will have patients who have resisted progression, yet continue. This is like MSI-high in on checkpoint inhibitors. They're continuing to do great, and these modest progressions that you see are not pseudo-progression, at least that by the classic definition, but these patients are continuing to do great. And so I think that's what. And also, even in patients who come off therapies, when immunotherapy really works, that same benefit can actually extend to later lines of therapy. So that's one thing. The other part about pseudo-progression is really seen much more, I think, in melanoma, and to a lesser extent, I have seen it in MSI-high. I don't, you know, once again, I don't know whether or not there's a lot of pseudo-progression, but I think that I've certainly hoped for pseudo-progression in prior immunotherapy trials and found out, unfortunately, that it is progression. So I think that those are really... And pseudo-progression, when it happens, is early, and I think that difference, that discordance between PFS and OS, at least in MSI-high, is late, where they're just sort of imaging abnormalities, you know, but patients are doing great, and we keep them on. Yeah, and I'll just chip in. The reason we put PFS as the first secondary endpoint was because of concern we had about pseudo-progression. Mm-hmm. Again, this is challenging because we're first-in-class innovators in the space. You just have—There's no precedent data. What we did know is that pseudo-progression can be difficult to manage, if you, you know, end up with a lot of people who are labeled as progressors per RECIST 1.1, and that's why we decided to use ctDNA as the first endpoint. Now, we took on different risks there, and obviously, as it turned out, we actually got it upside down because we don't seem to see pseudo-progression. PFS does seem to be robust as a metric of efficacy in this, with this therapy, in this setting. And our short-term definition of ctDNA response did not work out, because of the unexpected changes in the control arm. So we've learned a lot, but obviously, this is innovation in action. But the good news is, pseudo-progression doesn't seem to be a big concern, and that makes life a lot easier. Great. Thank you both. The next question from Catherine, it's another question for you, Dr. Hecht. Absent any kind of- Yeah ... tail effect, what would you want to see in both OS and also PFS for GRANITE in order to characterize it as a significant breakthrough, rather than merely hitting the minimum HR of 0.8? I'm sorry, I missed one word. Could you repeat that first sentence? Yes. Absent any kind of tail effect, what would- Oh, tail effect, that was the word I missed. I'm an oncologist, so which is different. I would love to see a huge effect. That's maybe different than, as you pointed out, you know, the hazard ratio of 0.8, which is kind of, you know, the minimum, particularly in an overpower trial. I would love to see, you know, a hazard ratio that. And when we talk about when we look at curves at a meeting, and the curves, you can put fingers between on the Kaplan-Meier, you know. So I would say I would love to see, you know, below 0.7, and I'd love to see an OS benefit. The one thing I would say is that those metrics, and at least in colon cancer, those of us who've done phase III trials, we have submitted our data to an organization, a nonprofit called ARCAD, and we put them all together. But PFS, you know, those sorts of things are probably better for cytotoxic chemotherapies and perhaps for, you know, anti-angiogenics, where, you know, you don't have that huge tail of the curve. And what I would say is that, you know, we have to be creative, and if you have a group of really long-term beneficiaries people who benefit, I think that, you know, hazard ratio, you know, you may have a more modest hazard ratio, but still better than 0.8. I would certainly like significantly better 0.8. But if you also have a tail of the curve that looks really robust, it does take a while to get. I think those would be really interesting discussions with the agency. Great. Thank you, Dr. Hecht. So our next question comes from Roy Buchanan at Citizens JMP. Please go ahead, Roy. Hey, thanks for taking the questions, and the presentation was great. Just a quick one for Dr. Hecht, since Andrew mentioned it, that just maybe get your thoughts on how this technology might be best implemented, in the adjuvant or neoadjuvant setting in your practice. Thanks. I think that's a great question. And as I said earlier on, I mean, we had to choose something, and actually, we, meaning Gritstone, had to choose a place in order to, you know, in order to go. And I think that's really a drug development question. So adjuvant is after surgery. Neoadjuvant, of course, is before surgery, and we have ongoing discussions as to whether what would be the best place in order to look. I do believe that earlier is better. I also think that in MSI-high, we get wonderful responses in the preoperative setting, whether it's in colon cancer, like the NICHE-2 trial or in rectal cancer, and we'd love to be able to replicate that. So the downside of those trials from a drug development standpoint, not from a patient standpoint, from a drug development standpoint, is that it takes a lot longer, it takes a lot longer necessarily to get your endpoint because, you know, people are cured with surgery, and you're waiting for people, you know, to recur. And so I think that I think that an approach like this clearly, you know, I think is open to that, and I think that's one of the nice things, would be adjuvant or, or neoadjuvant. One problem with neoadjuvant might be, I could foresee, and I don't want to talk for Andrew, would be just making certain that things get up and running. In rectal cancer, we have a lot of time to get things up and running because people get all their therapy up front. Adjuvant is also really, really... I think all these are really, really, you know, attractive places to go. I'm happy to spend other people's money and talk about trial design because I'm really excited about that. Okay, great. Maybe- I don't know if that answered your question, but the answer is yes. Yes, yes, please look at all these things. Mm-hmm. Okay. Another kind of, I guess, speculative long-term-ish question. Just your thoughts on maybe other combinations that you'd like to see tested in this population. You mentioned KRAS mutations, about half the patients, and cell therapies haven't really worked so far, but other companies are combining vaccines and cell therapies. Is there anything you'd like to see in terms of additional combinations? Thanks. I think there are sort of two parts to that question. So one is that, you know. And for full disclosure, I am running cell therapy trials, and I would love for them to be positive because I think there's certainly space for all of these things to be, you know, you know, to, you know, to help our patients. But you're right, about 50% of patients have KRAS mutations. I'm not. Well, we are sort of on the cusp of anti-KRAS drugs. KRAS inhibitors are a great undruggable target. I don't think inhibitors are gonna lead to long-term benefit. And then I can sort of once again defer to Andrew, but I'm not, you know, actually, I'm not aware of there being a difference from your neoantigen and presentation standpoint about KRAS wild type versus KRAS mutant tumors. So, you know, I think through your EDGE technology, you're sort of looking at all of these and trying to see whether or not the key thing is whether or not they're expressed or not. As for other combinations, you're right. I mean, there are cell therapy is advancing faster than we can do the testing, if that makes sense. And so that the genetic engineering, you can engineer all sorts of things in cell therapy. You can put cytokines in, you can put antibodies in, you can put all sorts of things, but it's a place where we haven't seen a lot of benefit yet. And so I'm all for people running those trials, but I think that, you know, as an earlier question is, you know, contribution of, you know, components, if you're combining this with a cell therapy trial or a vaccine with a cell therapy trial, that becomes much more difficult. I'd love to see, you know, the longer-term benefits first. Got it. Thank you. Thanks, Roy. I think we have time for one last question, Tara. Yes. So our final question comes from Corinne Jenkins at Goldman Sachs. Please go ahead, Corinne. Good morning, guys. Maybe a couple clarification or follow-up questions from us. On the kind of risk stratification, I guess, how did you think about the choosing the median, variant allele frequency greater than 2%, and did you look at other cut points? What might they have shown? And then I think just a clarification here, but in terms of what portion of patients were censored, I don't know if that was, that was disclosed with the, with the high-risk patient population, so I'm curious how many of those patients were censored. In the high-risk population, there, there's still some censoring, as you saw from the curve, there are censoring marks. I don't have the exact figure to hand, but it was something like 35. ... 30-ish% censoring, something like that, 30%-35%, I think. So in other words, the majority of patients had had events, but not all. In regard to the first question, the standard approach, when you don't have obviously, a priori knowledge of what the midpoint would be, is that you just take the median, and then you cut either side, because obviously, then you've got the maximum number of patients in the control on either side of that, threshold. So that's how we derive the 2%, and then we just apply that across the board. We did look at alternatives, and the Hazard Ratio wanders around a little bit, but it was sort of generally in the 0.5-0.6 range. So again, I don't think those individual hazard ratios mean very much because these are immature data analyzed using a post-hoc analysis. It's a potential window into the future. What matters is the mature data that we'll have shortly. Okay, thanks. Great! Okay, thank you, Corinne. So with that, I think it's time to wrap up, and I apologize. We do have other questions from patients in the portal, but we don't have enough time to answer them. If you do have questions you'd like answered, please send them to ir@gritstone.com, and we will then be able to respond back to you. And I noticed there are some patient advocates in the portal, so obviously, I'm if your questions are important to you, then please do follow up and we will endeavor to get responses to you. I'd like to thank Howard, obviously, for joining us today, congratulate you on the amazing things you've done with the time that you've been blessed to be given. I think it's really wonderful what you're doing for the community, and thank you, and thanks for joining us today. Thank you, Randy, for all of your efforts, obviously, just cranking away, not giving up, because we obviously have a lot still to do for the, this disease and for these patients. Thank you to Tara, and the LifeSci Advisors for hosting our Webex today. And again, if you have questions, ir@gritstone.com. Otherwise, we look forward to seeing you quite soon with mature PFS data this quarter. Don't have an exact date yet, but stay tuned, because obviously, these data are potentially, groundbreaking and very important for the field. Of course, for the company, but most importantly, for the field and hopefully for patients. So with that, thank you. Have a wonderful day.