All right, so welcome to the Fireside Chat with Boundless Bio. My name is Michael Schmidt, Senior Biotechnology Analyst with Guggenheim. It's my great pleasure to welcome Zach Hornby, CEO of Boundless. Zach, welcome. Thanks for joining us. Thank you for hosting, Michael. And so again, before we jump into Q&A, could you just sort of set the stage and provide a quick overview of the company before we again go into some specifics? Yes. Boundless Bio is a next-generation precision oncology company that was formed in 2018 to address one of the largest remaining unmet needs in oncology, and specifically, this is patients with oncogene-amplified cancers, so for those who have followed the targeted therapy revolution over the past two decades, they would note that targeted therapies work very well for certain types of genetically defined tumors, namely patients with gene fusions or with point mutations. This is where the TKIs and monoclonals have great data, are approved, and generate billions of dollars of sales. However, there's another type of onco-driver that is even more frequent than both fusions and point mutations combined. That's patients with amplification, so instead of a mutation, they have a wild-type sequence, but too many copies. Now those exact same drugs that have been approved for fusions and point mutations don't work for patients with amplifications, are not approved, and therefore, patients with amplified tumors have fewer therapeutic options, statistically inferior survival, and just poorer outcomes. Our scientific founders discovered some of the underlying biology and root cause as to why patients with amplifications fare worse, and it comes down to a cancer-specific culprit called extrachromosomal DNA, or ecDNA. ecDNA are cancer-specific circular units of DNA that reside in the nuclei of cancer cells, but they independently replicate and transcribe and give rise to high copy number amplification, and so our company was formed to leverage this unique insight into ecDNA biology and to identify novel targets that could be pharmacologically exploited to deliver new therapeutic options for patients with amplified tumors. So super interesting. And so maybe just talk about the frequency of gene amplification as an oncogenic driver compared to mutations or translocations, which are, again, sort of addressed with more traditional therapies? Yeah. So mutations and fusions account for approximately 15% of all tumors. And they're particularly common in certain tumor types like lung cancer, colorectal cancer, sarcomas. Amplifications, on the other hand, are even more common. They account for about 25% of all tumors. So once again, more common than point mutations or fusions, yet no therapeutic options aside for HER2 inhibitors. That is the sole exception where there are approved drugs for amplifications for HER2-amplified tumors. Right. And so why can gene amplifications not be addressed with traditional TKIs or other small molecule inhibitors? Yeah. So what we found, and we've published this, in fact, just last week, we had three concurrent publications in Nature, including the cover story, is that when amplifications occur, there's basically two ways they can occur. They can either occur on the native chromosome, where the gene resides based on the human genome map. And in that case, it's called a tandem repeat. And it's a chromosomally locked alteration. But these ecDNA-based amplifications, this is where units of DNA break off the chromosome, circularize, and then begin to replicate and transcribe. It turns out that when that occurs, it leads to a lot of heterogeneity, such that each tumor cell has a slightly different genome than the cell next to it. And so when you have this bulk tumor population, every cell is genomically different, meaning they each have different fitness advantages under various conditions. And so when therapeutic pressure is applied, whether that's chemotherapy or targeted therapy, certain cells that are less fit will be susceptible and will die. But other cells with slightly different genome will be more fit and will continue to thrive, and those will become the dominant clones. And so it is this tumor heterogeneity enabled by the ecDNA that is causing cancer cells to become rapidly resistant to targeted therapies. Great, and yep, talk about how your platform that you've developed can sort of take advantage of this biology to discover new product candidates. Yep, so what we did at Boundless Bio is we created a custom-built platform that we call Spyglass that we started with the unmet need, the patients or the tumors that are driven by amplification. We have developed and exquisitely characterized hundreds of amplification-driven models, so hundreds of different cell lines representing different tumor tissue types and different oncogenic drivers. We've characterized them based on what's the specific driver, what's the copy number, and does it reside on ecDNA or on the chromosome. Then we basically are able to segment our library of models into either ecDNA positive or ecDNA negative. Then we apply various target identification screens to try to identify targets that are synthetically lethal only to the ecDNA positive models, but not to the ecDNA negative models. By running the screen in this fashion, we are able to identify targets that are specific to ecDNA, that are fundamental to the ecDNA biology. So I'll give you an example. If we have, let's say, a panel of 10 ecDNA positive lines and 10 ecDNA negative lines, we can run a whole genome CRISPR screen where we, one at a time, knock down each of the 30,000 encoding genes to identify which genes, when knocked down, kill the ecDNA positive cells but spare the ecDNA negative cells. That is how we identify targets that are unique, synthetically lethal to ecDNA. And then maybe just a follow-up. So we talked about gene amplification, and I think you mentioned 20% of cancers. 25% of it. 25, right. Yeah. How frequently is ecDNA the source of the oncogene amplification? Yeah. So for high copy number amplification, which is defined as eight or more copies, ecDNA accounts for 54% of tumor specimens. So just over half of patients with amplification will have ecDNA. Gotcha. And then, yeah, so you just talked about your discovery approach and platform. And so you've obviously identified a few programs that you're pursuing, including BBI-355, which is in clinic. And so, yeah, maybe just talk about the mechanistic rationale of your lead program, and then we'll talk about some of the trials that are ongoing. Absolutely. So it starts with those target ID screens that I talked about a few minutes ago. We run those in an unbiased or hypothesis-free way. So we just let targets bubble up empirically based on the data. But as those targets do self-identify, we map them into biological pathways so that we can understand the higher-order biology that's at play. And it helps us start to begin to understand how do these ecDNA form, how do they function. For instance, how do they replicate, how do they transcribe, how do they localize, how do they export. And as we start to gain that insight, that helps us understand where we might intervene in the life cycle of the ecDNA. And so currently, we are intervening in three different facets of the biology. Our lead program, which Michael just alluded to, is called BBI-355. This is a clinical stage CHK1 inhibitor. The biology here leverages the fact that when ecDNA are present in cancer cells, it leads to high replication stress. This actually was the subject of one of our Nature manuscripts last week, is that ecDNA causes high replication stress. Replication stress, if unmitigated, is lethal to cells. Replication stress will lead to mitotic catastrophe and apoptosis. What these cancer cells do in response when they have replication stress is they invoke the DNA damage repair pathway. The master regulator, the key orchestrator of the DNA damage repair pathway, is CHK1. These cancer cells that have ecDNA have replication stress, rely upon CHK1 for their survival. If we now inhibit CHK1, we disallow their self-regulation. It's like taking the brakes off of a sports car as it's careening around a corner. And so that does then lead to this synthetic lethal mitotic catastrophe and apoptosis in those cells. So that's the premise of our lead program. Right. And so CHK1 has been obviously identified in the past as a cancer target in general. And so any learnings from other CHK1 programs that have been pursued, and how could those apply to your program? Yes. So many pharma companies have worked on CHK1 in the past. And about a decade ago, most of the big companies like AZ, Merck, Pfizer, Genentech brought CHK1 inhibitors into the clinic. And occasionally, they did have single-agent activity. But in that era, there was no biomarker hypothesis, no patient selection. So the prior CHK1 inhibitors delivered sporadic activity. They were ultimately always dosed up to maximum tolerated dose, where they did encounter on-target hematological toxicity. So up until now, the risk-benefit ratio didn't justify phase III studies with the prior CHK1 inhibitors. When we identified CHK1 as a synthetic lethal target with ecDNA, and knowing that ecDNA would be the biomarker, we first profiled all of the prior clinical stage CHK1 inhibitors to see whether we Boundless might want to take one off the shelf from pharma and develop it now with a biomarker. However, as we profiled all of the prior compounds, we determined that all of them were suboptimal for our purposes. Either they were insufficiently potent, they were insufficiently selective, or they were not orally bioavailable. So what we decided to do was to discover and develop our own in-house compound that was really optimized for potency, selectivity, things like cardio profile, and then oral bioavailability. Great, and so BBI-355 has been in phase one now for some time, so what have learnings been so far from the study? Yeah. So BBI-355 entered the clinic last year. And when we IPO'd earlier this year, we shared the clinical data available at that time. And at that time, we were still the study itself is called the POTENTIATE study. It's got three parts. Part one is single-agent dose escalation. Part two is escalation of our agent in combo with certain targeted therapies. And part three is kind of like a phase II dose expansion efficacy. So at the time of the IPO in March, we updated the data that was available at that time from part one, where we showed that with oral every other day dosing, our compound was well tolerated and had dose-proportional exposure at 20 mg, 40 mg, 60 mg every other day. At 80 mg, we began to encounter the predicted on-target hematological toxicity. So that was declared to exceed the MTD. 60 mg was the MTD. We showed that at that dose level, we were in the exposure range associated with efficacy in our preclinical models. We announced that we were advancing into part two. We are currently in part two. Specifically, these are combination arms where we're addressing patients with EGFR amplifications. We give our drug BBI-355 in combination with a wild-type EGFR inhibitor, which is erlotinib. Our second arm of part two is looking at patients with FGFR amplifications, where we administer our drug in combination with an FGFR inhibitor, futibatinib, which is provided by Taiho. We have a third arm that hasn't opened yet, but will shortly, which is patients with CDK4 or CDK6 amplifications, where we will co-administer with abemaciclib provided by Lilly. Those combo cohorts are ongoing. We haven't shared data with them yet, other than to say that so far, there's been no emergent combination toxicity, and that we intend to update the data from those arms second half of next year. What drove the specific selection of those three amplifications? Yeah. A couple of things. First of all, just the unmet need. So EGFR, FGFR, CDK4/6 amplifications are relatively common. Each of them are frequently on ecDNA. And each of them have approved drugs for the target, yet those drugs haven't been approved for amplification of the target. So in other words, there was a commercially available inhibitor we could use, yet we knew that that inhibitor alone didn't deliver compelling efficacy. So the hurdle was relatively low in terms of if our combination works, we could hopefully quite clearly demonstrate superiority. Right. So what are you looking for in terms of some of the data next year coming out of those cohorts? Yeah. So the historical experience for all of the EGFR inhibitors in patients with EGFR amplification is a combined 13% response rate and very short durability. So we would certainly love to double that, so deliver probably a 25 or 30%+ response rate with reasonable durability, let's say four-plus months. Obviously, the higher, the better. This is a tissue-agnostic design, so it could be any tumors with amplifications. For FGFR, the data is quite similar. The FGFR inhibitors have delivered a collective 16% response rate. So there too, if we were to double that, we'd want to be in the 30%+ response rate range. And I think that's meaningful for two reasons. One is it would differentiate from what the targeted therapies have delivered on their own. But also, I think as we've all seen across the industry, typically drugs that have a 30%+ response rate in metastatic tumors generally lead to approval. Great. And then your second program, BBI-825, targets a different target. Again, remind us of the mechanism of action of that program. Yes. So BBI-825, which entered the clinic earlier this year, is what's called a ribonucleotide reductase inhibitor. And so that's RNR is the short term for ribonucleotide reductase. As it turns out, when cells are making DNA or repairing DNA or replicating DNA, they need the building blocks, which are called nucleotides. And there are two mechanisms by which cells can gather nucleotides. The most common mechanism is called the salvage pathway. They're just recycling nucleotides from other uses within the cell. This is the most energy-efficient pathway, and it's the pathway that's relied upon by healthy cells. Cancer cells have a higher dependency for nucleotides because they're generating and cycling through more DNA. And so the salvage pathway alone is insufficient to satiate cancer cells. So they rely on a second pathway, which is called the de novo synthesis pathway. And this is assembling nucleotides anew. The rate-limiting enzyme in the de novo synthesis pathway is RNR. And so what we found is that cancer cells with ecDNA have this high need for nucleotides. Therefore, they have this dependency on the de novo pathway. And therefore, if you inhibit the rate-limiting enzyme, you basically starve them of their metabolic needs. And once again, that is apoptotic in these cells. Yeah. And then talk about in which context this might be important and how have you designed your development strategy around that? Yeah. This is an important question. So the first program that we're talking about, BBI-355 and the POTENTIATE study, that was designed for patients whose tumor is driven by an oncogene amplification as its original sin or its truncal alteration. The 825 program in its STARmap study is going after a different setting. These are patients who actually start with a point mutation as their initial driver, specifically a MAP kinase point mutation like KRAS G12C or BRAF V600E. What's been shown in the clinic is that when patients with these point mutations are treated with target therapies like a G12C inhibitor or a V600E inhibitor in combination with an EGFR inhibitor, about 50% of these patients develop resistance that is mediated by oncogene amplification. These are the patients we're going after with the 825 program. It's patients who start with a point mutation but develop amplification as their primary resistance driver. Now we are trying to address that resistance. Any comments on the ongoing phase I study and the early takeaways? What we've disclosed publicly is that we are still in the part one single-agent dose escalation. So far, the compound has been generally well tolerated and has demonstrated oral bioavailability. And that's all we've disclosed so far, but we've projected a further disclosure the second half of next year. Okay. And I think you're the first RNR inhibitor in clinical development, if I'm not mistaken. We are currently the only selective RNR inhibitor in clinical development. One other company had previously also worked on this. Taiho Pharmaceutical also had initiated a phase I program but had discontinued it, and they were not developing in this context, not ecDNA, not MAP kinase, not amplifications. And so, new target. So, any expected on-target toxicities or dose-limiting toxicities as you think about the biology of the target? We had thought before we even ran our tox studies that perhaps there would be hematological toxicity. There are other approved agents who hit RNR as one of their polypharmacology. So as an example, the compound gemcitabine does a lot of things. One of the things it does is inhibit RNR, and it has hematological toxicity. So we had thought that might be possible. In our preclinical studies, our GLP tox studies, we did not encounter any hematological toxicity. In fact, the compound was very well tolerated, both species, all doses, no mortality. So there really wasn't a lot observed preclinically. So it's a bit difficult to say what might ultimately be the dose-limiting toxicity in the clinic. Right. And then, yeah, you talked about some of the mechanistic hypotheses. Can you talk about the sort of phase I study and perhaps expansion cohorts that are being planned? Yes. And what are you looking for to demonstrate in those? Yeah. So the phase one study, again, it's called the STARmap study. It has a similar design to what I described as the POTENTIATE study for our 355 program. And so what it means is it's a three-part design. Part one is your classic single-agent dose escalation with BOIN in terms of statistical criteria for escalating. Part two is combination escalations. And the patients either are G12C colorectal, in which case we're combining with a G12C inhibitor and cetuximab, or they're V600E-driven colorectal, in which case we're combining with encorafenib plus cetuximab. And then once we complete the part two combo dose escalation, we would proceed into part three, which is the efficacy expansion cohort at the recommended dose of the combination. And then I know you've also talked about disclosing early data next year from that study. So what should we expect there? The anticipation for the data disclosure next year would be certainly the part one single-agent safety, PK, et cetera, and then hopefully some data from part two and possibly early part three, where we would hope to see some signs of clinical activity. Great. And then I know you've also been working on a diagnostic, the ECHO diagnostic that could help in development of these drugs and in terms of detecting ecDNA. And so can you talk about, yeah, what you're doing there in terms of the DX development? Yes. So currently, with current clinical assays, ecDNA is not part of what gets read out by Guardant or Caris or Foundation. So we, Boundless, have had to develop the capability to detect ecDNA. And one of our key strategic imperatives was we did not want to disrupt patient or pathology workflow. We wanted to leverage how patients are currently being treated in the developed world. So we set out to design an assay that would leverage standard next-gen sequencing data. Because what's most common for patients is they get one of these large onco panels like Foundation Medicine, where 300-400 genes get sequenced. And so what our scientists were able to devise was a software algorithm that takes the standard data that comes off an NGS assay and simply recalculates those data in a unique way, looking for the fingerprint of ecDNA. We have developed that. We partnered with a diagnostic company called SOPHiA GENETICS to add in all the quality control and do all the independent validation and verification. We recently presented that validation data at the ESMO conference. The FDA granted us what's called non-significant risk status. This assay has now been IRB approved and is being utilized in our ongoing POTENTIATE clinical study. Great. Well, thank you. I think it's time to wrap up. So appreciate the update, Zach. And thank you so much for your time. Thank you, Michael. Appreciate it.