ADC Development Revolution: Leveraging Translational Models for Next-Generation Cancer Therapeutics

Hello, everybody, and welcome to Crowncast. This is the podcast from Crown Bioscience that aims to evaluate the trends in oncology and drug development. My name is Benjamin Wilkin. I am a product marketing manager at Crown Bioscience. I, work on the ADC portfolio, and my background, started in GMP manufacturing and operations, and then I transitioned into organ on a chip, organoid, and three d bioprinting over the last couple of years, and that brought me over to Crown, where they, you know, specialize in translational models. Today, I'm joined by Davy O. Yang. He is the SVP of innovation, at Crown Bioscience, and he has significant experience in the ADC and biological space, driving first in class ADC and biological programs, with 20 IND enabling molecules, 10 approvals, 10 IND approvals, for small molecules and biologics, and 44 peer reviewed publications. Davy, welcome aboard onto the podcast. And, how are you doing today? Doing good. Thank you, Ben. Cool. Davy, you want us to start off by telling us a little bit about a little bit about... More about yourself and what you've been up to for the last couple of years. Sure, Ben. So, yeah, my name is Davey Ouyang. So I'm serve... Serving as SVP of innovation at Chrome. So actually, I joined Chrome in May 2025. So just about five months I've been with Chrome. But I'm no stranger to Chrome's because I worked for Crohn before from 2015 to 2021. Actually, back then, leading Crohn's in vivo pharmacology operation as well as r and d. And so in between last four and half years where I was with Crown, actually, was back in drug discovery space, worked for biotech and startups. So actually a lot of program also around ABC innovations. And in my earlier career, worked for pharma companies as well. So I've been working with pharma biotech startups on the drug discovery side, but also sales like doing service for drug r and d. So we're happy to have these opportunities and talking about ADC therapeutics. It's certainly a very exciting area with Ben. That's great. Thank you. And so speaking of, you know, talk... We're here today. We're we're talking about ADCs. Would you be able to give us kind of a brief overview about what ADCs are, and why they've become such a focus for oncology drug development in in the recent years? Sure. So we all know ADC stands for antibody drug conjugates. Right? So actually, I really like the analogy of describing ADC as biological we saw against cancer. Right? So the antibody really serves as a navigation system where it seeks out and binds those specific tumor associated antigens on the tumor cells. But you have the linker payload, which can form a well controlled warhead. Right. So which release cytotoxic agents where it's lifted and minimize the damage to the normal tissues. So really over the past decades, have seen ADC transformed from a very niche biologic to a mainstream therapeutics, especially following the approval of EnHER2. A lot of ADC is targeting various TAs being approved, including those like connecting four, chop two tissue factor, 40 receptor alpha, and more recently the same at. So in the especially in the solid tumor space, it's really an excitement to see so many ADCs got approved and really start to benefit cancer patients. At the same time, you know, ADC is also become powerful growth engine for a lot of us start up for biotechs. We have seen record breaking all licensing and partnership deals, which reflect the confidence of a large pharma in this modality. So it's really a very exciting space. But having said that, so not every ADC succeeds. We have also seen, you know, over 10 billions of failures in ADC program globally. And those failures can be due to the wake target validation, you know, discovery and the translational understanding gaps, as well as other reasons. So that's why data driven preclinical modeling and deep target biology insights become very critical. And that's exactly where Chronbio's integrated platform can make a real difference there. Given the complexity of these these ADCs, with their need for precise targeting, effective payload delivery, and manageable, toxicity, how important are models for testing these, and is that where it kind of models things like organoids and PDX, other translational models, come into play? Sure. That's a very good question, Ben. So if you use any word to describe ADC, certainly complexity is the right word, right? So first of all, there are like, you know, target and this is biology is there like, you know, so we have seen now more than 60 targets ADC have been tested in clinic. So every different target behave very differently across tumor types and that their expression level internalization mechanism, as well as tissue distribution are all different. That also determine how efficiently ADC drug can be delivered. And then you have a different antibody format there. Right? So you have a traditional mono specific monoclonal antibody and your bispecific. You have a smaller binders like VHH, like peptide binders. All those actually, you know, influence the binding affinities, abilities, internalization pathways, as well as how deep the binder can penetrate the tumor. Right? And then that also comes with a linker technology, linker chemistry that governs the stability of the ADC drug and also the payload release mechanism. So both that shape as efficacy and the safety aspect of the drug. And finally there's distinct the payload class that's come into play, such as topoisomers inhibitors, tubulin disruptors, as well as the DNA damaging agent. All those drive a unique MOAs as well as a resistant mechanisms. So it's very complex drug. So because of that reason that we use a model that refracts as a patient tumor become very critical. We certainly also offer two d and three d cell nine models to test as a ADC binding, internalization, as well as cytotoxicity, and also using the CDX models for the in vivo efficacy testing. Right? But I think what really differentiate Crum from others is our cross platform patient derived model system, meaning our organoid and paired PDX models. So now we have more than 800 of organoid models and about 2,700 of our PDX models. So there we can use the organelle platform for the highest throughput cytotoxicity screening of your ADC drug or pain nodes. And beyond potency assessment, we can also use organoids for deeper MLA studies, so as a high content imaging, right? We can understand the drug penetration, we can understand the payload response patterns and use that in a more of physiologically relevant system. And then we can extend that insights into our PDX platform right where we capture the tumor heterogeneity, know, stromal contents and also vasculature and also more importantly you have the in vivo pharmacokinetics plays into a role right so together you know I think this system form a very powerful translational bridge that connecting the functional biology and a systemic pharmacology there. And then beyond the models, we also, you know, can reinforce this bridge through our biomarker service where we have some immunohistochemistry assays, as well as multiplex immunofluorescence assays and special omics assays to profile the patient tumors as well as our models, including the argument and PDS. And there we can have a deeper understanding of a target expression and a special context to guide ADC innovation. So I see this as an integrated workflow that allow us to link ADC design parameters to the clinical relevant model platforms, really turns a complex drug engineering to a more predictable translational outcomes. The ability to to use models for each individual kind of strength and and not just having all these models and and being able to use each one, but to kind of connect the dots between the models is is very important. So how how do you optimize this ADC development when you're combating different diseases, and what are the key considerations when you're selecting some of these models like organoids, PDX models, or or both when designing some of these, studies for these ADCs? That's a very good question. So very... Also a very practical question. Right? I mean... But I see this is not about choosing one model platform or another. Right? But I think serves the complementary roles across ADC R and D process. I think, for example, Argonoid is especially powerful in the discovery stage when you're doing your lead optimization. Right? So, you know, our asset platform allow you to do a rapid screening of a multiple antibody linker payload combinations and to provide you that insight into potency, pedal release rate, and allow you to rank your candidates. Right? And then when you narrow down to a couple of leads, PDX or other in vivo models can become essential, then very... It can allow you to start a PKPD relationship, biodistribution, and also potential toxicity there. So you know, and then which is crucial to your final candidate decision on selecting your drug candidates. Right? So that's in the discovery stage. And then in the translational stage, I think both platform also plays a unique role. Where you can see organo platform is more like a population study in vitro, which can reveal different response across tumor types and also to uncover the predictive biomarker beyond just a target expression and pedal since data through there. And then PDX can extend this approach to the in vivo system showing how micro environments, you know, vasculature system exposure of your ADC, how that improves efficacy as well as a toxicity profile. And then we can even enhance this framework through the biomarker profiling of the model right to ensure that target expression pattern in our models closely mirror your patient tumor that you are target. So in addition to that, we also have, you know, organoid that is gene edited engineers, or the clinically portrayed PDX model to enable you to explore the resistant mechanism of ADC drug or to the other combination therapies. So altogether, I think our paired organoid and PDX platform and then target validation capabilities really provide you a very complete translational tool kits. It's perfectly suited for every phase of ADC drug discovery and and the development from very early on to translational research. And discussing more about kind of the some of the these translational models, what are the current limitations or challenges when using organoids or PDX models for development? And how are researchers working to overcome some of these? Yeah, that's again a very good question. I think there are certainly, as long as I can see, two different type of challenges, right? It's either scientific challenges or operational challenges, right? So at Chrome, we certainly have the biggest commercially available PDX library that will allow us to run PDX trial study, which we call mass clinical trial to mirror real patient diversities there, right? So we have a great number of PDX model we can put in use. However, in the organoids, our connection, although already 800 models, but it's a number game, right? I mean, as the bigger your library is, the better you can find models reflect as a large, diverse patient population. So we're still looking to expand our organoid connection. Another challenge on organoid side is it's a throughput or the operational challenge, right? A lot of our organoid actually the growth rate is quite slow. And then also you need to revive, expand, then set up your acid plate. So even a simple cytotoxicity assay sometimes can take two, three months to complete. So that post the timeline challenge for a lot of drug developers. So at Crown, now we have optimized the organoid operation by selecting about 200 of more robust, fast growing models and make some acid ready. Now we have about 200 models there. So by doing this, that will allow us to deliver reliable screening results within six weeks. So that largely shorten the timeline and also facilitate the decision making there. And there are also certain other challenges like what you can see, both systems still limited in immune context, right? So complicating the combination studies, but I see those combo studies more of a proof of concept type of study rather than serving the real translational purpose. And then also there's data integration and biomarker modeling challenge will require strong bioinformatics infrastructure there. Luckily, Chrome has a very capable data science and bioinformatics team can help you do that. And lastly, I will say certainly it's always a challenge to to get that clinical relevant late stage resistant model connection, right? So I think that also truly mirrors a real world ADC patient population, because when the ADC enter clinic, the first patient population will be facing is actually the next stage patient, whereas the expression profile of your target, the response to various payloads can be very different from the treatment naive patient. So Crohn is actually now using a lot of resource, try to build up those heavily treated resistant models. Yeah, that's awesome. So a couple of things I know on the marketing side, you know, the pretreated resistant models are something that people are always interested in in more information on and I think that you you nailed on the head with the the the fact that many of the the newer drugs that are gonna get tested are going to be on on those patients that have already passed through a certain, you know, you know, first line, second line treatment. So that's really good for for people who are in the development space. The second thing is is six weeks on the the organoids. You know, for anybody who's handled organoids, I got my start handling organoids, almost ten years ago at this point and and scaled up operations. Six weeks is impressive from time to start off a study through through results, once you consider acquiring tissue and culturing and expanding and then finally running the assays. So awesome awesome work on, things that... You know, ways to address some of these key issues for researchers. Finally, we we have one final question for you, Davey. So with the advances in in personalized medicine, how do you see the role of patient derived models evolving in guiding ADC design for specific patient subgroups? Great question. Will say something also a million dollar question, right? So as you see more and more of oncology therapeutics moving towards a precision medicine space, right? And that requires a preclinical testing also need to be precise, I mean, we're patient derived model not only can evaluate efficacy, also serves as a tool to predict the patient response as well as to guide the clinical strategy there. So there's a concept become more and more popular which is called co clinical trial or co clinical programs, right? Where you can align the preclinical data with ongoing trials and then to refine, there are several things we can do to refine the biomarker hypothesis and to optimize enrollment criterias and also to identify your responders early on, right. And also more importantly, like you know to expand the indication more rationally. So there's several things can happen there if you run the co clinical trial. And together with that I think another aspect is by integrating biomarker assays, I'd say proteomic, special omics, all those assets, we can actually map out the targets distribution, all the immune context in a way that each tumor revealed the different each of model revealed different subgroups of patients, right, which can mostly likely to benefit from specific ADC designs. So the ultimate goal there is really to combine this patient derived model systems. Organoid and PDX and to, you know, incorporate target validation capabilities that Chrome has to allow us to firstly design smarter ADC drugs, right, to enable that, but also to enable us to design a smarter trial. So that way we can make the development of ADC drug more predictive, efficient, and also personalized. Awesome. Thank you so much. So, Davy, thank you so much for your time today. I want to give one more chance for you if there's any final thoughts that you have for the listeners. And then also, how can people reach you if they wanna, you know, discuss more about ADC development and Crown's capabilities? Absolutely. So Crown, we really see ourselves not just a CRO, but we try to become your discovery and transformational research partners. We can combine, you know, our advanced model system, you know, our biomarker analysis assets, deep understanding of data science and the bioinformatics to make not just ADC drug, but more broadly other therapeutics as well. This R and D process is more efficient, predictable and also clinically meaningful. So hope we can help you there and to facilitate your R and D process. And you want to reach me, you can send me an email. My email address is davy. Ouyangchromebio dot com. Happy to discuss any future project with you. Excellent. Yeah. And for all the listeners, definitely, definitely reach out to Dave if you have any kind of ADC questions. He's got a lot of experience in this in this area, so definitely take advantage of him if you can. And thank you all for listening. Hope that this podcast has been meaningful for you, and that you've learned a little bit more about the the the translational models and the, direction and importance of ADCs. So thank you.