iwNHL 2023 | New therapeutic targets in DLBCL

Ash Alizadeh: Hi there, I’m Ash Alizadeh from Stanford University and with my colleagues, Dr Laura Pasqualucci from Columbia and Dr. Ken Young from Duke. We had an exciting session to hear about three very different topics. We heard first from Dr Laura Pasqualucci about co-operativity between chromatin regulatory factors and chromatin modifying mutations in large cell lymphoma and follicular lymphoma. That was quite an interesting body of work that you’ve summarized for us. Also heard from Dr Young about new ways to think about therapeutic targeting of large cell lymphoma looking at the landscape of somatic mutations and new opportunities. I shared some ways we might look at both of those types of phenomenon using cell-free DNA to see what may be going on when we try to tackle these therapeutic vulnerabilities. I first had a question for you, Laura. One of the things I was intrigued by was should we be looking at these genetic alterations in these factors like CREBBP and KMT2D through a lens outside of histones and chromatin? In particular, thinking about the difference between lymphoma types for where the mutations occur, do you think there is anything that we may not be aware of for activity of these proteins on non-histone proteins? Any opportunity to see if there are new functions for the mutations, particularly in CREBBP?

Laura Pasqualucci: Well, I think that we certainly should look at that. We know very little as of now in the context of lymphoma, but both for CREBBP and KMT2D, also more recently we know there are non-histone proteins that are modified. I agree totally with you that there may be differential functions that could be related to these proteins in the different subtypes of DLBCL. Unfortunately, I don’t have much to say, but proteomic analysis, taking advantage of the models we have in our hands, many of us, would shed light on that. The KMT2D acetylation by CREBBP is an example that certainly this thing is happening.

Ash Alizadeh: Right. But do you think that the key, the central role is still histones, that’s the way in which these things are perturbed to deregulate gene expression through histones?

Laura Pasqualucci: Well, I think histones is important because clearly you do see a correlation between the expression of these genes and the localization of the proteins. But the other protein factors that may be recruited to these loci together with CREBBP or KMT2D or even the compensatory mechanisms that could come into play when CREBBP or KMT2D are lost will have an equally important role. In fact, that’s one area we’re trying to pursue as probably many others.

Ash Alizadeh: Great. Ken, you told us about very different ways of thinking about therapeutic vulnerabilities in large cell lymphoma, looking beyond, even within, specific hotspot mutations. A number of different ways of thinking about it, among those, if you had one to pick as the top of your list in terms of prioritizing, which are you most excited about of those strategies? Can you tell us a little bit?

Ken Young: Yeah, thank you. I think the most appealing to me is the P53 pathway. I think that this is people are already working on for 50 years. All attempts from pharmaceutical company, academic centers failed in the past 50 years efforts, including Roche that spent million, billions of dollars. I think this is probably the first priority because this is present in any type of human cancers. If we are able to make one single breakthrough going to be applicable in any cancer type, even individual mutations happening. Those are probably the first things that we’re going to moving forward and we’re looking for advice and opinion from all the experts, and looking for partners for collaboration in this area and moving into the clinical investigations.

Ash Alizadeh: Great. Now not all the folks that are watching had a chance to see your talk. Can you tell us how you would approach tackling P53 – as what is the unique way in which you’re tackling it in lymphomas?

Ken Young: You have three different approaches. One is artificial intelligence which is driven smaller molecules screening. We already screen 20 million compounds. We already find three compounds, they are very effective, able to inhibit tumor cells carrying these particular mutations. The second thing I feel is that now we are moving into immunotherapy centuries. I think the bispecific antibody targeting P53 mutations engaged with normal surrounding T-cells probably could be more even powerful. Now, certainly combination could be, to think about it. The third one is vaccine. We’re developing a particular injection, intramuscular injection, use of vaccine and they’re able to achieve quite a reasonable response. So I think these three approaches, but particularly antibody-based could be very powerful tools that in the future to think about it the same we follow for myeloma patients, for the DLBCL patients, bispecific antibody treatment. FDA approved five bispecific antibodies targeting myeloma and lymphoma. I think for P53-based bispecific antibody treatment could be a breakthrough consideration in the future.

Ash Alizadeh: Great, great. Laura, you mentioned the importance of these mutations in the key chromatin modifying genes as early events in the development of lymphomas and as kind of ways of defining the progenitors of – what’s your vision for how that knowledge would have us pursue therapeutic vulnerabilities. You talked about ways of playing one chromatin modifier off the other when there’s a deficit. Can you talk us through that process and how a therapeutic pathway development may work?

Laura Pasqualucci: First of all, I think therapeutic approaches may be different if you want to try and target the precursor cell, or if it’s a vulnerability that is maintained by the tumor cells and which therefore the tumor cell can remain addicted to. Because if we want to develop approaches to eliminate the precursor cell or maybe even to predict or prevent relapses when we can detect this mutation as you showed in patients when they are still healthy, that would be quite complex. Obviously, we would have to look for approaches that are not toxic and that would provide true advantage to patients.

If it’s a vulnerability that is maintained by the tumor cell as we showed for P300, I think that that obviously may, I think, would be worth it to explore. We don’t have drugs at the moment if we talk about P300 that are specific, as you know, and probably one would have to play with dosage. But there are already compounds so that though not being specific, have shown limited toxicity in clinical trials and, therefore, I think that that could be suggesting a therapeutic window. Obviously, there’s a lot of work to do, I think for identifying other vulnerabilities, and there’s a lot of efforts academic and companies. I hope again we’ll have a better answer next time. But maybe since this is meant as a conversation, I think this is a good opportunity for me to ask you instead, since you also gave this very, very informative and amazing talk as usual on the use of circulating tumor DNA for predictive and maybe even decision-making processes. I think I would like to know your answer to these related questions.

Ash Alizadeh: I showed data on applying circulating tumor DNA through the timeline of a treatment of a patient with aggressive lymphomas from diagnosis through early prediction of response through the course of therapy to measuring residual disease at the end of therapy. Of course, these tools show a lot of promise, but for lymphomas, I think the key questions that came up after the session are the big ones, which is how do we take this information and actually make a difference for patients. As some of the folks in the audience mentioned, there are now clinical studies being designed and executed for taking patients with specific types of alterations and giving them therapy. For instance, we have a risk-adapted trial in large cell lymphoma where patients who are not responding quickly get treatment escalation through the course of therapy, adding, for example, a bispecific to the backbone, logistically difficult to execute as an interim assessment.

There are now other ones being talked about. Some people asked about the end of therapy milestone and early CAR administration or early bispecific administration as T-cell engaging type therapies that would help clean up the leftovers after chemo-immunotherapy have not achieved a cure. Then the other point that was brought up is CAR as a dramatic therapy that we now rely on for our patients, doesn’t work in every patient. The treatment failures after CAR are more prevalent than the successes. So, how do we overcome those failures with ctDNA? The ability to measure it is the first step. The second, third, fourth steps are building a study to intervene, administering it successfully, getting a response, and then showing that you improve outcomes. Each of those steps is a big lift and we’re working on them. I guess the one challenge that we haven’t really addressed, and I think some related work in CLL and myeloma have shown us is that circulating – MRD, I should say – MRD can be different in the context of different subtypes of the disease and different therapies in the subtypes. It becomes kind of a little bit complex in terms of designing and executing a study. Just imagine if CREBBP mutant large cell versus EP300 large cell are as different as, say, mutated versus un-mutated CLL, for example, in their MRD responses. How do we design the clinical studies to address these questions? But I think we need to walk before we run and that process is just getting started.

Laura Pasqualucci: Why do you think P300 is in fact sort of counter-selected as a mutation in relapsed/refractory?

Ash Alizadeh: I was hoping you would tell me. I was hoping you would tell me. One, we don’t really understand – another pathway that we see as a pathway involved in B-cell homing to germinal centers that you may have seen where different members of that pathway are on two different sides of the volcano. We’re just trying to understand that with working with Jason Cyster, P2RY8 and S1PR2, are on opposite sides. One is selected for and one is selected against. We don’t totally understand why these things are happening. If you have good ideas, we can measure it, but we don’t necessarily know what it means.

Laura Pasqualucci: It’s certainly one thing that is emerging from these studies in general at any level is that the interaction between the genetics and how genetics instructs the environment with lesions that many genes that are important in fact for the delivery of signals from T-cell or from the microenvironment will be very important in the future, or currently, right.

Ken Young: This pathway you present is very interesting. Now, we are not only seeing in lymphoma, but we’re also seeing in myeloma, particularly in the myeloid neoplasm MDS/AML. Do you think it will be a different, they play different roles in the biology in oncogenesis or maybe in MRD testing or maybe drug targeting?

Ash Alizadeh: I guess, I think, in the myeloid malignancies we’ve learned that not all mutations are created equal for MRD detection. That the FLT3-ITD versus NPM1 versus t hat the gene matters for what you’re measuring. Often when using tools that don’t target the therapeutic vulnerability of the pathway. Right? It’s not BCR-ABL by ABL inhibitors. It’s not FLT3 targeted therapy. In lymphoma, I have to say that we’ve seen much less dramatic evidence of clonal selection at the level of mutations and pathways surprisingly. I was wishing we would see more, but at the moment, I think linear rather than branched evolution, at least in large cell lymphoma, seems to be the rule… with the branching being an exception or the fraction of the genome that’s branched is small. But maybe things will be different as we do more of these things over time. And maybe it’s because of the things you showed that the types of mutations that we’re tracking probably reflect very early events. Aberrant somatic hyper mutation happens as B-cells become B-cells and become crazy B-cells in the germinal center as early events. If we have a hold on that, it’s unlikely that changed dramatically through the course of treatment.

Ken Young: I see. Many physicians still have questions now. Now, MRD testing has become a routine standard practice for myeloma and leukemia, but not well accepted by physician on lymphoma. How are we going to position MRD testing in the future for those patients and how are the healthcare systems going to support this kind of testing?

Ash Alizadeh: It seems like lymphoma’s stuck somewhere between the solid tumors which have already moved on and there’s now clinical, and actually active use, of ctDNA, for example, in colon cancer after colon resection for choice of adjuvant therapy decision-making, and the leukemias and myeloma as you said, where we, I think, have been very married to our PET scans and seeing the disease on a scan. A lot of that, of course PET scans have been transformative and super important, I don’t mean to take anything away from it, but I think we have used images whereas leukemia and myeloma were more used to treating numbers and cells that you could see under a microscope. When CR rates started to go up, it was natural that you try to use more sensitive ways of measuring it. I think we need to move towards that. I think we just had a symposium organized by the Lymphoma Research Foundation around ways of getting there. I think there is a path, but unfortunately a slow path.

Ken Young: We have to merge easily with the solid tumors while FD-LC-MS already have proof of five types of solid tumor fully covered, but the lymphoma left out. Why?

Ash Alizadeh: Actually, diffuse large B-cell lymphoma last summer was – we protested and they did cover. They cover for CMS, not in FDA, and so diffuse large B-cell lymphoma MRD testing by clonoSEQ can be paid for. But the challenge is what you do with that information right now, and it’s mostly a prognostic discussion with patients.

Laura Pasqualucci: Probably standardization will be critical when one thing sets, extending this worldwide.

Ken Young: Also, you see the myeloma already have MRD-driven clinicals trial going. Is that going to be a duplicate in the lymphoma in the future? I think that might be worth consideration into design, really a qualified patient population.

Ash Alizadeh: I think patient selection based on MRD and lymphomas is already pretty far underway; using MRD as an endpoint like in leukemias and myeloma is emerging.But I don’t know the lymphoma clinical trial that has had MRD as a true endpoint, a regulatory endpoint for approval. For myeloma, a couple are going and CLL has a couple. Great. Any closing thoughts about the session and what your takeaways are, Laura?

Laura Pasqualucci: Well, continue to invest significantly in basic research, both applications and knowledge of fundamental mechanisms that we all know are important to develop and push the field forward.

Ken Young: I completely agree, Laura. I think today and Riccardo’s presentation, I think are really inspiring me. We should be thinking about regulatory non-coding genomic components and how that impacts cancer and lymphoma pathogenesis. They may become important diagnostic, predictive and therapeutic targets. This area we have never explored in the past. Now when all the technology comes up, I think this area, we should think about it now.

Ash Alizadeh: I agree. I think seeing the mix of things, the diverse topics that we talked about, I agree with you, Laura, that it’s basic research that makes these things possible and from basic research to application of basic research to translation and meetings like this to bring together folks who work at the basic level to translation to the treatment of the disease, so I think it’s a bright day for lymphoma and meetings like this. Thank you, iwNHL, for organizing this.

Thank you.