Pioneering Strategies to Unlocking the Potential of KRAS

Hello, everyone, and welcome to CrownCast, the podcast from Crown Bioscience that aims to discuss the emerging trends and viewpoints in oncology, drug discovery and development. I'm your host for today, Michelle Dawn Mooney. Today my guest is Dr. Rajendra Kumari, Crown's Executive Director of Integrated Solutions. Dr. Kumari has been with Crown over the last 12 years, but her background in oncology and preclinical research spans more than 20. Over the years, Dr. Kumari has held a number of leadership roles, including chief operating officer, chief scientific officer, and General Manager. Around six years ago, she moved into a more global role as Global Head of Scientific Communications. And for the last two years, she has served as Crown Bioscience's Executive Director of Integrated Solutions. Dr. Kumari, thank you so much for joining us today. Thank you for the introduction. Great to have you here again. So today's topic is KRAS, which is an area that is very close to your heart and one where you have deep expertise. So to start us off, could you tell us a little bit more about your background beyond the brief history that we just heard and how you came to be in the role you're in today? Yeah, so it's been, I've had some great opportunities in my career to establish a business unit and company in preclinical oncology. So I gained a lot of experience firsthand working with a range of different companies, including, you know, small companies, startups, academics, larger companies, biotech and pharmaceutical, and also being able to work with a range of different therapeutic modalities, different targets, and using different models as well. I think that's always nice to be able to do. I think there's a number of different advances that have happened over the year. And again, it's been great to be part of those as well. I think actually bringing these types of experiences and models to a wider community has really been one of the primary objectives here and part of Crown, we can do that as well. So in the different roles I've had within Crown, I worked in operations, so as General Manager and Chief Scientific Officer for Crown Bioscience UK. And again, working with local clients and giving them outreach to Crown's greater capabilities across the globe. Then moving into R&D, helping with the research and development of new models. Again, there's lots of challenges and lots of changes in the field. And Crown again is able to help facilitate research by creating new models. And then in this more recent role, the integrated solutions, again, looking at the different capabilities that are needed to really help drug development. It's a great chance to look at how we can position new technology, new models in the drug discovery sort of journey, as well as identify where things are needed, and also develop those alongside our client needs. And ultimately, this is going to be benefiting patients. So there's always an ultimate goal to this. And so it's really has been a great chance and opportunity to work in this ever evolving field. And thank you for sharing that. So the oncology field moves very quickly, as you know, not just in terms of science but also in the technologies and models we use. But focusing on KRAS specifically, can you tell us what KRAS is and why it's such an important target in oncology research? Yeah, I mean, out of all the targets, I think KRAS is one of the ones that's been around for a very long time. There's been a lot of research in the area. Actually, you know, KRAS alongside HRAS was one of the first cancer genes that was identified. And then amongst all the kind of RAS genes is the one that's most frequently mutated. And it makes it one of the most frequently mutated in cancer just generally. So this is the reason why it's so important in cancer. So it really does drive progression, promotes uncontrolled cell growth, survival, and also resistance to therapies. If you look back at the normal function of KRAS, it really kind of acts as a first line sensor almost, you can call it a sensor if you like, because it can be temporarily activated by growth factors and or tyrosine kinases. So growth factors like EGFR, and what it does, it actually transduces the signals from that cell surface to the nucleus, so via different cascade of different signaling pathways, which is really complicated. But just to simplify, I mean, such as PI3, ACT, mTOR, you got RAF, MEK and ERK. And then it influences key processes. So an arrange of different cellular processes and key for oncology is obviously that sort of cell differentiation growth, there's chemotaxis and apoptosis as well as other mechanisms. And it's actually a really small protein. And I think for something so small and being so important, it just shows how so critical in these sort of pathways. So it's a part of the sort of GTPase family and it binds to GTP in its activated state, and then in GDP in its inactivated state. So it acts like a switch and toggles between the two. And that's why it acts as a sensor. And the mutations, so this is the normal function. So when in cancer, there's a mutation, there's actually a gain of function and activation of those downstream signaling. And it causes oncogene addictions, the cancer cells become very dependent on that KRAS signaling to make it, you know, to make that cancer sort of grow. And it really does drive the growth and progression of disease. And because it's so highly frequent, obviously, it's impacting a number of different sort of cancer types. So it's a key regulator. I think that's what we're trying to say. And the other thing that's really interesting about this is those mutations can be very tissue specific. And the frequencies vary. And I think this is again, something that's very interesting. So like for example, in pancreatic cancer, a high unmet need medically, 90% have KRAS mutation, that's a phenomenal amount. Whereas other cancers like colorectal cancer, it's probably about 40% in non small cell lung cancer, it's probably about 30% percent. And again, those are, you know, big cancer types, where there's a lot of interest in developing new agents. And again, if you look and if you break it down, so I'm just going to break it down into the different sort of sections and subtypes, there's different types of mutations. And these are somatic mutations that can rise as a result of those single amino acid changes, those substitutions. And typically you'll get something like a G12 and then substitution. And that is the biggest one. So there's probably about 81% of cancers that have that mutation. You've got mutation for G13, lot less about 14-15% percent. And then there's others, there's Q61, and that's again a very small proportion. So you can see that G12 mutations are really very much more frequent. But again, they do vary across those cancer types. So you'll see G12D mutation, which is very frequent in those pancreatic cancers. And again, it makes up like almost 40%. And then again, in colorectal cancer, it's very common. Whereas in lung cancer, you see more G12C. So there's a huge sort of distribution of those mutation types. And that's just down to the sort of biology of the cancer as well. And the association as to how that maybe that cancer arose. So for example, the G12C in lung cancer, maybe it could be strongly associated with the smoking related DNA damage. Whereas in pancreatic cancer, the G12T mutations may rise because of that sort of chronic inflammation. So there's different environmental factors that may drive those types of mutations. So it's very fascinating to hear and learn about how this spread means that different tumors respond differently, maybe to those mutational influences. And it's fascinating to hear how different tumors respond differently depending on the mutation. You mentioned somatic mutations. So just to clarify for listeners, are these KRAS mutations genetically inherited or are they influenced by environmental exposure, or maybe changes that occur over the course of one's life? Yes. So these are sort of the latter, sort of they happen because of circumstance. And like I said, and that's why you get that sort of different tissue distribution as well. So, and then associated with that, obviously you get poor prognosis. So again, that depends on the sort of type of cancer, the type of mutation. And then obviously the sort of type of lifestyle that there is for that patient. KRAS has historically been described as an undruggable target with decades of research dedicated to trying to crack it. Can you tell us about the evolution of KRAS research? What characteristics of KRAS made it so difficult to target for so long? Yeah, and again, I think it's undruggable for such a long time. And it's like, oh, then why bother? You know, why is it so important to keep trying to hit this target? Obviously, yes, it's because it is a key regulator. And if you look at some of the sort of types of the role that these mutations play, you can see why it's important to continue to try and make it druggable. So the KRAS mutant tumors, for example, they're often very resistant to chemotherapy, or to other targeted agents. So for example, in lung cancer with EGFR inhibitors are ineffective in KRAS mutated tumors. And there's also Microsatellite Instability in some of the KRAS mutated colorectal cancers, which limits some of the immune checkpoint inhibitor benefits. And then you have other mutations. So this is not just about KRAS, there's also other mutations that are occurring. So co-mutations like STK11 in cancer, in lung cancer has been shown to be associated with resistance again to immunotherapies, and also other mutations like TP53 has also a worse prognosis. So looking at these different sort of mechanisms, you know, are really important. It's not just alone with KRAS, but we have to hit that KRAS first, before we can potentially solve some of the other problems. And as you know, we've learned more, we've been able to profile cancers more with identifying different mechanisms as well that potentially could be play a part in KRAS as a result of KRAS as well. Historically, it's been classified as undruggable. And that has changed, as you know, but the reason why it was undruggable was partly due to the way that the type of protein it is, it's a very small protein, small protein structure. So the tertiary sort of structure of RAS really shows that there's a lack of deep hydrophobic pockets, on that sort of smooth surface, which then limits the opportunities for designing and making drugs that are highly potent and also very selective. So these drugs need to be very selective. So because of that, it's been difficult to actually design drugs. Then there's also the issue that that whole sort of toggling between GTP and GDP has also caused the issues because there's a real high affinity and really this is like picomolar affinity binding to GTP, is bound in the active state. So once that happens in these mutated forms, it was believed to be very difficult to interfere with that. So again, that's been very challenging. And also there's very high cellular GTP concentrations, which again means it's hard to kind of get in the pocket, let alone trying to knock off GTP to kind of bind to those active sites. So, yeah, very undruggable for a very long time. And because of that, again, we must recognize that there was a lot of effort to still try and hit that pathway. So again, this KRAS influences so much biology that downstream there's obviously targets that we can also trying to hit to, you know, with different types of inhibitors. So again, some targets were also upstream that were being targeted with different sort of drugs and monoclonal antibodies. And so the key pathway here is obviously the Raf/MEK kinase. But again, many of these agents showed limited activity in these KRAS mutated cancers. And that's because there's a dynamic feedback mechanism with a lot of these signaling pathways. So even, and KRAS is a key regulator for that. And then also there's normal, this sort of tissue toxicity with some of these other targets. So no matter how we tried to target the pathway, it still came back down to trying to hit KRAS. So now after almost like 40 years, and that's a long time, right? The discovery of an inhibitor, that has actually been approved, which was in 2021. And this was specific to A mutations. Remember all the different types of mutations we just talked about, and the different type of cancer types. So this is only for G12C. And in 2021, was for non small cell lung cancer. Obviously more recently, that's changed as well. And so these cancers need to have that G12C mutation. And the reason for that is in the design of the drug. And this is where it's been really smart in how scientists have been able to, you know, really tackle this problem. And the breakthrough came in 2013. And so this is again, using a library of compounds and looking for alternative ways of maybe tackling that little protein that's causing us some problem. So the screen was, the intent was to actually to see if there was any way to, you know, identify inhibitors that may actually covalently bind to the cysteine residue in that G12C, so very specific. And this led to the development of the first compound, so it's referred to as compound 12. And that actually binds to a pocket within the switch-II region of KRAS. Now this was previously unrecognized, and this is again how these drugs are really giving us insights. So doing that research and investing all that time, we wouldn't have learned all this about that KRAS, but we would have just assumed it was still undruggable. So again, I think that's, you know, strides that have been made in, you know, making these specific inhibitors. Sometimes these inhibitors become tool compounds, they don't always become drugs just because of the different types of properties. And that's really what happened with compound 12. It had low potency, but it was still, they were still able to show that it could inhibit the activity of that mutated KRAS. And then eventually after different sort of iterations and designs, the first sort of compound that was developed was ARS-1620. And it had enhanced potency, it had good bioavailability in vivo. And it was tested with models like CDXs and PDXs. And so these are cell line models and patient derived xenografts. What they also learned, which again, think is really interesting, and again, it just shows how, you know, we don't know everything we have to learn as we are making these drugs and then approaching these different diseases, is that it was able to bind to the inactive form, so the GDP bound form. So it can maintain it in that sort of inactive state. And again, that wasn't something that was known. And because of that toggling between GTP-GDP, what they learned is actually even in the active state, it was still happening. So there was an opportunity for these inhibitors to get in and hold it in that GDP position. So again, forty years of research, we're still learning about this little protein. So I'm curious to know, was that breakthrough enabled because KRAS has a lower affinity for GDP or because binding pockets are more accessible when KRAS is in that GDP bound state? It's binding to the cysteine, which is, you know, not, it's very specific, obviously, to that particular sort of mutation. And I think it's just the affinity for both GTP-GDP is very high. But it's just interfering, I suppose, with switching back out again, which is why that sort of specific domain is quite, you know, interesting that they didn't really know that it could be possible by binding to that specific region. Even now with second generation inhibitors, you're seeing some can now also bind with the GTP bound form as well. So again, it's just the way that sort of compounds are designed and how some have a different sort of ability to bind to different parts of that protein. We've seen incredible drug breakthroughs come over the last few years, but how are current challenges around drug resistance being addressed? Is combination therapy a major focus? And then what does prognosis look like for patients undergoing treatment for KRAS mutant cancers? Yeah, no, I think these small molecule inhibitors, inevitably, there's always resistance that's going to emerge in the clinic. And if you think about the number of agents that are in the clinic, it's again, only two. So again, we're still learning a lot. And you know, with the first agent sotorasib, this entered clinical trials in 2018. You know, and there was very good indication from the, I think it was a code breaker trials showed favorable outcomes and limited toxicity, which again was really great news. So they did receive accelerated approval. So again, you can see how if a compound is really good, it goes through for approval very quickly. And that was for non small cell lung cancer. And then afterwards in the following year, there was another trial with adagrasib, again, very similar sort of mechanism of action to sotorasib. And again, similar sort of responses in non small cell lung cancer. And again, these are patients who received prior treatments. So they're not like naive or anything like that. They've actually relapsed. These are advanced, locally advanced or metastatic disease as well. So really hard cancers to treat anyway. And there was, you know, good response seen, benefit seen with just these monotherapies. But there are some cancers where it wasn't very effective. And so understanding that why it wasn't effective was key to the next round of agents or the indications. So for example, I think I mentioned that in colorectal cancer with the G12C mutations, you would expect these agents to potentially have a good efficacy, but they didn't, whereas in non small cell lung cancer, it was very good. So it was a case of understanding why that was happening. And here it was because it was linked to the EGFR activation. So the rationale behind how to approach this was obviously to do combinations with EGFR monoclonal antibodies. And actually, this was very effective. And you saw same agents, different EGFR inhibitors, antibodies being approved. So just last year, the FDA granted accelerated approval again, for adagrasib in combination with cetuximab for those G12C advanced metastatic colorectal cancer. And then just in January, we saw sotorasib also have the same approval in combination with panitumumab, which is again, another EGFR monoclonal antibody. So again, you can see understand the resistance in a different cancer type, and then looking for a strategy to overcome it has really kind of paid off. And then, you know, there's other new agents that are coming through the clinic now as new KRAS inhibitors. And some of them have similar mechanism of action, some of them are slightly different, like I mentioned. So they may bind in the GTP form rather than just the GDP form. You've also got different new types of second generation inhibitors, which actually don't even bind up maybe something different. You've got pan KRAS inhibitors or hitting multiple KRAS mutations, and also pan RAS inhibitors as well. And that is because being so specific for a G12C mutation, that means that you're only hitting a small population. So there's probably about five percent and it's not massive. So by increasing the different mutations you can potentially hit, you're more likely to benefit more patients. So that definitely is a strategy. And also because of the resistance that's emerging to those two G12C inhibitors, it's part of it is due to other mutations being present in the cancer. So it makes sense to have a different approach here. So yeah, so those resistance is, you know, really key now. And the other thing is we're still learning about them, right? They've only just hit the clinic. And understanding the mechanism again, will take quite a bit of time to figure out what's going on. And all patients are different. So the molecular profile of these patients is really going to vary and understanding why one patient is responding versus another is the challenge. So I think that's where things like, you know, the combinations are really gonna important here. I want to dive a little deeper here. Can you walk us through some of the key mechanisms of resistance that researchers and clinicians are seeing today? There's a lot of information that is based on obviously tissue coming back from the clinic, being able to then model it in the preclinical stage as well, because obviously with all these new agents coming through, and the potential agents to combine it with, because you may be trying to hit another pathway, actually becomes a little bit of minefield and very sort of complicated. And then I think also as part of this is that sometimes there may be primary or intrinsic resistance. There could be acquired resistance where, you know, the tumors adapted and gained other mutations, or other sort of adaptive types of resistance, which is associated with non target related mechanisms. And we don't really fully understand some of their intrinsic resistance. But yet, you know, we do actually have some ideas and trying to identify how to pick those patients out before giving them treatment unnecessarily is also something that's key here as well in the research stages. Whereas the treatment that's sort of induced by patients being given the inhibitor, that is, I mean, it's quite fast, it's about four to six months after the initial response. And so it doesn't really give much time. And like I said, it can be caused by co mutations in other genes. And we've seen that in some of the clinical studies. So getting that information back and turning it around quickly just to extend that four to six months is going to be really important to those patients. And what they found from these studies as well, that it can actually be quite complex. It could be some patients just have one of the mutations, some patients don't have any of the mutations, might be an adaptive mechanism. Or it can be a mixture of different factors that are occurring simultaneously in single patients. So again, unless we research and investigate those identified biomarkers really key here to have a real integrated approach to this, not just think about how to look at the efficacy, but also how to maybe predict which patients are going to respond. We need to employ clever technologies, to kind of understand the multi omics side of our cancers. So, so we were learning. So we know that there are other mutations, specifically around KRAS. So that could be for the G12C, obviously there could be a G12D. There's also a Y96C, there's R68S, and these could be in that, you know, located in the switch-II pocket. So again, it stops the inhibitor binding. So in those situations, you can't really do much apart from look at alternative drugs. And so your sort of strategy will be slightly different to maybe an activating mutation in a downstream pathway. So the BRAF MAPK/ERK, MET and RET receptors upstream as well can all have a role to play in here as well. So you can then look at a strategy, maybe hitting it with something else to kind of bring down the overall response. And, you know, again, with the sort of these adaptive resistance, it's can sometimes be non genetic. I think that's the other thing that we, are learning as well. And therefore, there could be like an amplification, it could be due to the tumor microenvironment. And this is where immunotherapy then obviously, has a role because we know that KRAS actually has a role of in manipulating the tumor microenvironment protecting itself. So therefore, attacking the tumor environment might be an approach or strategy that's needed for certain types of tumors and certain types of patients. So yeah, different mechanisms that we kind of better need to understand. Bringing the conversation back to preclinical oncology. What are the best model options available today for KRAS research? Are some models more suitable than others? And then what are the key advantages or maybe limitations researchers should be keeping in mind? Yeah, no, absolutely. Those preclinical tools are really important, especially when you've got these novel agents coming through. I mean, obviously some of these are first in class, were first in class. Now we're looking at second generation, the best in class, looking at targeting resistance. And then there might be those combinations with other agents that have got nothing to do with the target itself. So having, and so like you said, it's multifaceted, and that's key here. And so therefore having a preclinical toolbox with a range of different models really can provide different options to evaluate those different therapeutics, those combinations, and also the difference of disease phenotype, because I think that is what we're seeing here, that not everything is going to be going to fit into one bucket. So therefore, we probably need to answer specific questions for the type of approach that's being taken for that type of cancer. So for example, most of the preclinical models can be quite simple. And I don't think there's anything wrong with that. I mean, the initial sort of journey that we've seen with the first in class inhibitors, they've actually seen, they've actually used those basic sort of cell line models, and then the xenografts. And then for more translational studies, they've used PDX models of patient derived xenografts. And these then become the sort of standard models that are used for other types of inhibitors, because they want to use the benchmarking data generating these models, so that they can compare new agents or the combination. So I think there's almost like a sort of a set sort of number of models that are used. But when you get a more complicated picture or there, you know, there's other sort of factors that need to be considered. Actually one, I mean, typically one model is not going to give you the full picture. And actually sometimes a handful of models isn't going to give you, you know, it's not sufficient, it's not a good range to answer all those questions. So you need to kind of pick the right models that are suitable to answer those kinds of specific questions. And the cell line models are very versatile. They can be used for efficacy drug screening in early discovery. They're easily engineered. So if you want to express a specific type of mutation, you can do that. You can use them in more complex systems like in 3D cultures or in those xenograft systems. And then like I said, for the more translational studies, you've got the PDX. And the reason why these are really important came about in the sort of last two decades. And I remember when, you know, being having developed PDX models, how this, there was a sort of ramp up in the need to represent different cancer types, but also have a large number from patients to develop a range, a portfolio of different PDX models. So the reason why they became so hot was because they really are more reflective of the patient tumors. They have the original tumor histopathology, the marker features, the heterogeneity. And so when you've got large models, they're actually very representative of the diversity of patients that you're probably using in clinical trials. And because of that, they're actually much better predicting pharmacologically what sort of response you're going to see in the clinic in comparison to some of those 2D models, those xenografts, which have very much adapted to grow in plastic. So these became, you know, really popular and replaced a lot of the cell line models. But then if you look now then at other models, and also technology that has become available, which can, some of this can actually provide deeper insights, will help maybe differentiate the treatment strategies, and also the precision sort of medicine approaches. I think that's really important. And it does actually link back to some of the limitations of those standard cell line models and the PDX models. And in fact, because of those limitations, know, sort of the limitations of CDX, because of the lack of clinical diversity, that's really where, like I said, PDX really excelled. So you can see how it changed. And then immunotherapeutics became really popular. And because the xenografts are immune deficient models, typically, you saw a shift into other types of models, which are more homographs, so where they have a fully functional immune system. So again, the types of therapeutics of the need from the therapeutics of development and the clinical need, really drives how these models are developed and prioritized and taken forward. And now because of the sort of combinations, we want to, we have to be clever and try and find models that actually can answer questions around multiple targets, some of which may be on the tumor, some of them won't be in the tumor microenvironment. Now, I mean, if you look at the PDX models, like I said, highly valuable, because they are clinically predictive. But in the situation of KRAS, they really aren't reflective of what's happening in the clinic today. So we have to be constantly evolving and adapting our preclinical models so that we can answer the questions that are coming out of the clinic. Now, know, at Crown, we're very fortunate that we have access to clinical centers through partnerships, so that we can create new PDX models. And we've been able to generate these from non small cell lung cancer patients who've been treated with either sotorasib or adagrasib and obviously their disease is progressing, which is why we're getting the tumors now, obviously, if they were responding, there's no tumor to get. So we actually receive these and can make PDX models. And the molecular profiling and the pharmacological testing of these unique and rare PDXs is key then to understanding why those patients aren't responding. So for example, if it's intrinsic, was acquired or adaptive, those are questions that are really challenging to answer, but we need to kind of try and do as much as we can with these models that they're highly valuable. And then trying to identify what treatment strategy can be used to treat the tumors. Again, it could be anything from either trying to recover the sensitivity to those KRAS inhibitors, or it could be a combination approach with a different target. And so this, so you know, we're trying to aim to increase the number of these pretreated models, because we can see a resistance emerging and we need to understand the full repertoire of what's happening in the clinic. So it does, you know, it does need, we do need to evolve. And I think that's what we're trying to do here as well. It's been great to talk about KRAS. I think it's a very exciting area. And after after four years, the undruggable has become druggable. And that's a significant step in tackling cancer. But it also, as hopefully, I've kind of alluded to opens up a multitude of different opportunities to find better or alternative strategies to target those specific tumors with KRAS mutations, really to help our patients receive the best treatment. And we know that the monotherapies aren't producing the results we want to see. So the combination strategy is really where this is going to be moving forward. And there's going to be obstacles in terms of trying to model those combinations. So very exciting times. And I think it's really nice to be in a position where we can help in that. We have a range of different models, super excited about the fact that we can actually look at immuno-therapeutics in combination with KRAS inhibitors. And again, having access and the ability to do humanized systems, having a range of immune competent models with specific KRAS mutations, and the ability to do engineering really does enable us to have a better understanding. And also, there's late stage disease. I think this is the thing, the late stage disease, metastatic disease, we see a lot of brain lesions in some of these patients, you know, being able to model those means that you need sophisticated technology. So I think we do always need to look at how we can advance these areas of research, not just with our models, but also technology really plays a big part in this as well. So I think that's where the field is moving to, and this is going to really benefit the sort of preclinical space for KRAS inhibitors, which hopefully then can help better inform the sort of clinical strategies that are being employed. Yeah. Dr. Kumari, thank you again. That was incredibly insightful. Appreciate your time here today. And it's clear that KRAS has moved from being considered undruggable to becoming one of the most exciting and fast moving areas in oncology research. I want to thank everyone for tuning in to CrownCast today. If you found this episode valuable, please feel free to subscribe on YouTube, Spotify or through the Crown Bioscience website at crownbio.com. I'm your host, Michelle Dawn Mooney. We hope to connect with you on another episode of CrownCast soon.