Stephen Ansell:

Welcome, my name is Steve Ansell and this is an opportunity for me to learn and enjoy a conversation with two of my colleagues. This is a focus on modern diagnostics in diffuse large B-cell lymphoma. As I said, I’m Steve Ansell, I’m at Mayo Clinic in the Division of Hematology, joined by two colleagues that are very expert and it’s going to be a fun conversation. Dr Rebecca King, who is a hematopathologist at Mayo Clinic. Dr Paolo Strati, who is at MD Anderson. I’m going to ask both of you just to take a minute and give a little bit more of a formal introduction to exactly your role. So, Becky, I’m going to turn it to you first. 

Rebecca King:

Sure. Thanks, Steve. So, as Steve said, my name is Becky King. I’m a hematopathologist at Mayo Clinic in Rochester, Minnesota. And I focus primarily on lymph node pathology and lymphoma with a specific interest, both clinically and academically, in aggressive B-cell lymphomas. And I’m involved in a variety of sort of organizations where I interface with both clinical hematologists treating lymphoma as well as hematopathologists who specialize in lymphoma. 

Stephen Ansell:

Becky, that makes you the absolute perfect person to be on the call today. Thank you very much. Paolo, I’ll turn it over to you. 

Paolo Strati:

Yes. Hello to everybody. My name is Paolo Strati. I’m currently a clinical and translational investigator in the lymphoma and myeloma department at the University of Texas MD Anderson Cancer Center in Houston, Texas. And I focus on immunotherapy and cellular therapies for B-cell non-Hodgkin lymphoma, mainly large B-cell lymphoma, follicular lymphoma, and marginal cell lymphoma. And from a translational perspective, my interest is mainly in trying to understand better the impact of the innate immunity on B-cell non-Hodgkin lymphoma. And I’m very excited to be here today. 

Stephen Ansell:

Paolo, you again make here a perfect kind of person to facilitate and contribute to our conversation today. So again, we’re going to focus in on large B-cell lymphoma. And Becky, I’m going to ask you to start us off here. Things have changed a lot. You know, when I started, and that was like a million years ago, you know, it was back in the working formulation days and a lot of things changed from then to now. Maybe you just want to give us your kind of overview. If you get a lymph node sent to you and you look under the microscope and you go, I think this is large cell, kind of how do you think about that just as you’re developing the diagnosis, just so we can understand kind of what buckets you’re thinking of as you’re doing the analysis? 

Rebecca King:

Yeah, absolutely. Thanks, Steve. So yeah, as you said, I think, you know, diffuse large B-cell lymphoma, which is sort of in a way the prototype for these aggressive B-cell lymphomas, used to be pretty easy from a pathologist standpoint. Fairly easy to recognize on the slide. You see, you know, replacement of the lymph node or other tissue by these large cells. And, you know, you do a B-cell marker, R2, you know, a CD20, and great, you’re done. You know, diffuse large B-cell lymphoma. So pretty straightforward. But as you all know, you know, this has evolved substantially. And so usually it’s less about, you know, what is the diagnosis from that standpoint and more about kind of what other information do we need to further subclassify this large B-cell lymphoma. There are, I don’t know how many, but, you know, more than 20 probably subsets of diffuse large B-cell lymphoma, some of which, of course, are specified based on the site of presentation. So a good example would be primary central nervous system large B-cell lymphoma. But the majority of the large B-cell lymphomas that, you know, we encounter in day-to-day diagnostic practice and you encounter are going to kind of fall into this category of, you know, diffuse large B-cell lymphoma NOS. Usually nodal presentation, but can be extranodal, can involve basically any, you know, I might get a biopsy from basically any site in the body. And so typically diagnostically at that point, you know, not only am I confirming the lineage, which is, you know, I’m seeing these large cells and now I know their B-cells, usually CD20, CD19, Pax5 positivity. And then, you know, the sort of standard of care right now is to establish cell of origin, which we know, you know, was originally described 25-ish years ago now based on gene expression profiling and, you know, described the activated B-cell subtype and the germinal center B-cell subtype of diffuse large B-cell lymphoma. Certainly 25 years ago, we did not have the capability to do that in standard clinical practice. And so, of course, that, you know, has translated over the years into immunohistochemistry algorithms. And now the most common one we perform, and that’s performed really around the world, is the Hans algorithm, which involves just three immunohistochemical stains, CD10, BCL6, and MUM1. And we assess those to basically assign a DLBCL into germinal center B-cell or non-germinal center B-cell type. It’s not a perfect recapitulation of the gene expression profiling, which we can get into, you know, more detail of, but it does give some information and, you know, has, as you guys can discuss more, you know, held up prognostically in many studies. So typically we do that. We also often will do, you know, MYC and BCL2 staining. It can help us in certain differential diagnoses, such as, you know, with things that may be Burkitt lymphoma, and gives us some indication of the expression of those proteins, although it’s not a substitute or a good indicator of whether there are rearrangements of those respective genes. And so what we do in all of our cases of aggressive B-cell lymphoma and what’s really recommended in standard of care now is to do FISH, to look for MYC rearrangements, and then as well as BCL2 and potentially, depending on your perspective and center, BCL6 rearrangements as well. And that’s to exclude or identify the cases that are what we today would call high-grade B-cell lymphoma with MYC and BCL-2 rearrangements or high-grade B-cell lymphoma with MYC and BCL-6 rearrangements. And I won’t speak to the, you know, clinical implications of that because that’s what you guys are here for. But typically, that would be our workup in DLBCL. So let’s say all that FISH is negative and I’ve got, you know, my Hans algorithm says, okay, this case expresses CD10. It’s a GCB. So then I make a diagnosis of diffuse large B-cell lymphoma, you know, GCB by the Hans algorithm. That’s kind of the, I think, brief nutshell. You can give obviously more details and tangents, but I’ll let you… 

Stephen Ansell:

We will do that in just a second, but thank you. That’s a great start. So, Paolo, I mean, I think, again, as I, like you, I’m often in the clinic and I take a look at my list of patients and I see somebody coming, oh, and it’s going to be diffuse large B-cell lymphoma. Give us your thoughts about, even before you got there, what’s your overarching kind of way in which you think of large cell lymphoma? 

Paolo Strati:

Yeah, I’m going to share with you a very clinical and real-world type of approach, so not mentioning anything experimental this time. And I tend to divide usually what I’m looking to when patients come to my clinic or, you know, even before they do it, based on whether they are previously untreated or in the relapsed/refactory setting, because the biological features that I’m looking to tend to be quite different. So for patients who are previously untreated, what I ask my, you know, fellows in clinic to make sure to report are, first of all, the histology. So DLBCL as compared to high-grade B-cell lymphoma, primary mediastinal, transformed follicular lymphoma, and of course, the less common subtypes, like Becky mentioned, Burkitt, for example, because typically, they will help us to understand whether we’re going for a regular intensity type of frontline, anthracycline-based chemo, like R-CHOP or Pola-R-CHP, or a higher intensity, like dose-adjusted R-EPOCH. As a next step, I do look into the cell of origin, as Becky mentioned, either by nanostring or gene expression profiling, actually, we don’t have it available at MD Anderson, so we still rely a lot on immunohistochemistry and the Hans algorithm. Though at this time, I have to say cell origin in my practice doesn’t have per se a therapeutic implication yet in the frontline setting. But it’s mostly for me to understand how much I should be worried for the next element, which is the double-hit status. Because double-hit tends to be present mainly in GCB as compared to non-GCB. So I definitely look into double-hit by FISH, and that will have an impact on my treatment selection. It’s typically for double-hit lymphoma. We will always go whenever possible for a higher intensity chemo. And then, of course, I look into stage IPI and CNS IPI. Now, when it comes to relapsed large B-cell lymphoma, all of the information becomes less relevant to me. In addition to patient-specific features like age, comorbid health conditions, or the timing of relapse and things on that line, in a biological perspective, what I end up looking into is really just a special of targets for therapeutic options that I have available. So I look into CD19, CD20 expression, CD30, and I will utilize either immunohistochemistry or flow cytometry and hope for the best. I mentioned at the beginning, I will not touch on anything experimental this time, but we start looking increasingly into new antigens such as CD70, BCL6, but far you mentioned it, anything that would open clinical trial options. And finally, what I’m realizing as we have more and more for relapsed large B-cell lymphoma availability of agents that target all of those, I see actually the morphology of lymphoma changing as these patients become more and more unnegative with no expression of any of any of those antigens and actually their morphology becomes more plasmoblastic and that becomes a real headache in clinical practice. 

Stephen Ansell:

Oh, that’s very helpful. Becky, you know, we’ve touched on a few topics as we’ve gone through here. We talk about DLBCL-NOS and we’ve spoken about GCB/ABC and then we’ve spoken about high grade and maybe double hits, triple hits. And then we’ve spoken about, I put it in the beginning, dark zone. Can you just talk us quickly through again, that spectrum, what’s what and how do you land in each of those buckets? 

Rebecca King:

Yeah. So this is definitely an area, Steve, that’s, I think, evolving and constantly evolving. And, you know, we’ve been, you know, since we had a classification and a WHO classification in 2008, a WHO revised classification in 2016. And then in 2022, you know, now we’re kind of in this area where we have these two pathology classifications, the international consensus classification and the WHO-5. And I say that as an introduction just because there have been changes in this area in each of these kind of updates because this is a rapidly evolving field. So, you know, originally what pathologists and clinicians recognize is that there are cases that look on the slide, you know, somewhere intermediate between DLBCL and Burkitt lymphoma. We know Burkitt are, you know, the most aggressive cases, right? But things that sort of don’t quite fit into either bucket. And as time went on, we learned that a lot of those cases, but not all, have what we’ve referred to as the double hits, the MYC rearrangement and the BCL2 rearrangement, particularly. And so in, you know, 2016, the entity of high-grade B-cell lymphoma with, you know, MYC and BCL2 rearrangements and MYC and BCL6 rearrangements was introduced to kind of capture that more aggressive biology and allow the clinicians to identify those patients who may need more intensive therapy. You know, that has evolved with, you know, more variable data on the cases that have BCL6 and MYC rearrangements and evolved in recent years with the discovery of what’s called the dark zone signature. So that is a gene expression profile signature that is found in dark zone B-cells, so normal B-cells that are in the germinal center, the dark zone germinal center, it’s a very proliferative spot of the body. You’re trying to refine your B-cell repertoire. So you have rapidly dividing and turning over cells there. And so this dark zone gene expression profile has been described in Burkitt lymphoma, as well as these MYC and BCL2 rearranged double hits, as well as a subset of diffuse large B-cell lymphoma that don’t have those features, that don’t have the MYC and BCL2 rearrangement by FISH. So, you know, this is all still- I don’t want to say preliminary because it’s been pretty well established now that this is true- but it hasn’t really translated, you know, maybe one exception, into routine clinical practice. But the, you know, these signatures that can be identified identify a substantial proportion of patients who otherwise would not have been considered high risk. They would have looked like they’re mostly GCB cases, as Paolo pointed out. And it’s identifying this subset of patients that would otherwise have been thought to have a fairly good prognosis and favorable biology that actually, you know, have a risk of doing very poorly. And so that is sort of, I think, the most interesting sort of step we need to make in terms of introducing that into the clinical practice. 

Stephen Ansell:

Very helpful. Paolo, you made such an interesting comment that I want to get your clarification on. So you said, you know, in frontline, first-line therapy, you’re kind of underwhelmed by the relevance of all this cool stuff that Becky’s sharing that has developed over the last few number of years. So does any of this impact your thinking about a choice of treatment? Do you do something different? For example, if you go, oh, this is more high grade or this is double-hit, do you do something different if you think this is ABC versus GCB? Or at the end of the day, it’s like, oh, well, we’re still doing R-CHOP like we did back in the 80s.

Paolo Strati:

Yeah. Thank you for asking for clarification. So we do utilize some of the information that Becky and other pathologists will provide. As I mentioned before, as for histology, we definitely know that high-intensity chemo tends to be associated with better outcomes for subtypes like Burkitt lymphoma, high-grade B-cell lymphoma, double-hit, and primary mediastinal B-cell lymphoma. You can cure a significant fraction of these patients, whereas we may R-CHOP to those who don’t have those histologies. As for the cell of origin, again, we don’t do yet at this time, at least in my institution, utilizing standard treatments, any different therapeutic decisions based on the cell of origin. But now one point that I wanted to touch on is that it would really be nice to understand matter who those patients are, who don’t have double-hit FISH, but as Becky mentioned, may have the dark zone signature. Just to elaborate a little bit more on that, while we do know that the dark zone signature exists and identifies patients who are also patients who are double-hit negative by FISH, we don’t really utilize much the gene expression signature in clinical practice. And there’ve been multiple attempts by multiple hematopathology groups to develop some sort of Hans algorithm for the signature. Understanding that every time you transition from gene expression profiling to immunohistochemistry, you’re losing a little bit, you know, 20 to 30% of patients probably who do have the dark zone signature will not be identified through an immunohistochemistry-based algorithm. It will still have a clinical impact because for those patients, there are two things that we could do. Of course, we can think on the line of double-hit lymphoma and maybe utilize high-intensity chemo, but there have been some retrospective translational data from the POLARIX trial, the randomized Phase III trial, based on which we can now utilize in the frontline setting, Polatuzumab vedotin, an anti-CD79b antibody drug conjugate in combination with R-CHP as compared to R-CHOP for patients who have an IPL2 or higher. And what our retrospective translational analysis has demonstrated is that actually the addition of polatuzumab to R-CHP seemed to be able to address this clinical unmet need, which are the dark zone signature patients. So I already, for example, in my practice, for patients who have double-hit lymphoma, where I feel like due to age or comorbid health conditions, it would be really concerning to utilize high-intensity chemo like dose-adjusted R-EPOCH, utilize Pola-R-CHP instead. And I’m going to start to do that for those who don’t have double-hit lymphoma but do have the dark zone signature if I had a way to identify with a simple technique such as immunohistochemistry the patient population. 

Stephen Ansell:

So you make a few important points, and Becky, I’m going to come back to you. And, you know, 25 years of going around and around about the gene expression profiling and then sort of molecular clusters. And we’ve heard a lot of things, originally ABC, GCB, and now multiple clusters. Where is that? What have we learned from that? Why have we not brought it to practice? We used the word Hans algorithm like six times because that’s what’s happening, and you mentioned right in the beginning, Becky, that it’s only partially really kind of predictive of which category people should be in. What are your thoughts about the molecular clustering and why has that not really got to real world really at this point? 

Rebecca King:

Yeah, so great question, Steve. So there’s many layers to this. I think, you know, in general, research science tends to be so far ahead, you know, to where we are clinically. And I mean that in the pathology lab as well as in, you know, in the hematology clinic. And part of that is the translation of the sort of practical of how to translate a test that’s developed in, you know, a research lab, just practically into a clinical setting for, you know, financial reasons, practical reasons, reproducibility reasons across different, you know, laboratories and institutions. You know, you referred to the molecular clusters, which are these fairly reproducible clusters that multiple groups have identified as true biologic subgroups of diffuse large B-cell lymphoma. And I think there’s a few reasons why it hasn’t sort of jumped into the routine clinic yet. And a big one is, well, which classifier, as we call them, are we going to use? Which set of clusters? They’re roughly identical, but not completely identical. And so, you know, is it just going to be that the Mayo Clinic decides which set of biologic clusters they’re going to use, and then MD Anderson decides which they’re going to use, and then a patient may move around and get sort of different assignments depending. So that’s one issue. And then another issue, again, is simply the practical one. Most of these studies are based on, you know, really comprehensive genomic data, both, you know, whole exome sequencing, whole genome sequencing, as well as copy number assessment, things that aren’t just routinely done in most labs, even at big centers. And so, you know, how much data do you need to actually, even though these classifiers are publicly available, how much data do you need to enter in there to get an appropriate output? And, you know, is your assay that you developed at one place the same as another assay developed at another place? And so there’s a lot of technical and logistical problems. And then another big one is, you know, is there clinical demand? Because what we do in hemepath and in test development should be and appropriately is generally reflective of what is needed in the clinical setting. And so in a way, I push that back to both of you to say, you know, why aren’t you demanding this, if you will, in your patients? I know, you know, it’s being done in clinical trials. But, you know, for, you know, Mr. Smith, who comes in with DLBCL, generally, I’m not hearing that you care, you know, which of these buckets he’s in. So I think that’s sort of an overview of that question. 

Stephen Ansell:

Well, that’s such a good point. And thank you for pushing it back to Paolo and me here. But Paolo, I mean, you kind of touched on the fact that actually you may not care that much. And so I guess I want to ask you your thoughts. You know, I think there’s a body of group of people out there that are saying, hey, if you have an ABC subtype, you really, that’s the group that, for example, you mentioned Pola-R-CHP, polatuzumab vedotin plus R-CHP chemotherapy, they may benefit best almost from that subgroup, provided the IPI is higher. And then there are some other folks out there who are going, hey, the recent data that showed tafasitamab plus lenalidomide-R-CHOP seem to have shown benefit across the board, but maybe actually that’s where I should give the GCB people that treatment. And then there are other folks who are going, well, don’t forget, as you touched on already, if there’s any sense of the double-hits or the, you know, the dark zone signature, you need to intensify therapy and you need to be giving dose-adjusted R-EPOCH. So should we be going back to Becky and pounding the desk and going, where is my dark zone signature? Where is my GCB, where is my MCD subtype molecular study, which by the way everybody says is bad, why are you not giving me a test that says that? What do you think?

Paolo Strati:

So that thank you for allowing me to clarify. So when I mentioned before that you know I don’t care much in this sometimes. I did kind of say that, but I preface by saying in real-world practice, and this may be different, of course, in a scenario where you have access also to trial options in the front-line setting. You mentioned among many, you know, tafa-R-CHOP, which is not yet formally approved by the FDA on the front-line. But for Pola-R-CHP, the FDA label is just based on IPI or two or above, just to clarify for those who may be less familiar with the literature. There have been some retrospective analyses, the most cited from Dr. Ashley Zadeh in the Journal of Medicine a couple of years ago or longer, showing that probably the benefit of Pola-R-CHP over R-CHOP was mainly driven by the non-GCB or ABC subgroup. And so that’s an open debate where some of us tend to be very conservative, and they will only go according to FDA label and they will pick Pola-R-CHP over R-CHOP every time the IPI or two or more is hit. Doesn’t matter what’s the cell of origin. And some others, including myself, I’ll say, tend to look into Pola-R-CHP as an option over R-CHOP in the frontline setting, mainly based on the cell of origin along with the IPI, of course, and maybe reserved to the GCB subtypes only if the IPI is really high, three or more. And hence, I would be more concerned about an aggressive biology. So again, in clinical practice, that’s the current implication. But back to your point, and again, to also provide some background to those who may be less familiar, there are currently many signatures out there based on very large data sets and very sophisticated assays, including single-cell RNA sequencing, gene expression profiling. But of all of these, only one has been made slightly more accessible to the LymphGen algorithm where utilizing a few gene mutations, we can identify the six clusters overall. However, as of today, having that piece of information about the gene cluster based on the gene mutation that many of us try to collect based on whatever customized gene panel is available at our institution, it’s more to accumulate information for future analysis or for selection of clinical trials. And there are actually currently in the US and Europe ongoing clinical trials trying to understand how we can utilize, for example, the cluster information to pick the best, let’s say, companion for R-CHOP plus X. But now I think the complicating factor, and once again, I feel like I’m moving from real-world clinical practice to a more clinical research type of approach. The two complicating factors are A, that we recently learned, for example, from work from one of my colleagues, Dr Michael Green, that in addition to the lymphoma clone, also the microenvironment matters. In that case, the analysis was mostly mentioned on a focus on response to CAR-T, so second line or beyond. But this may also be relevant to the selection of the right frontline setting. So we need to stay tuned as they may be implemented with this more like clone, focus, gene signatures. But also the extra complication is that there is a chance that we may be moving away from chemo. And everything we mentioned today in the frontline setting, both in terms of like simple features such as histology, cell of origin, double-hit, and then the more complicated signatures, they’ve actually been trained and validated in chemotherapy-treated patients, and it’s unclear what their clinical and biological relevance will be in the not so far future, where in the front line, we may utilize a purely biological or immunotherapy-based regimens. So we have to keep in mind these potential layers of complications as we have these discussions. 

Stephen Ansell:

Well, I mean, that’s exactly right. That’s the fun part of this conversation today. So maybe transitioning a little bit, and Paolo, you touched on this before about sometimes how you like to know about novel targets, and maybe the novel targets are on the tumor cell, maybe the novel targets are in the microenvironment. I’ll say right here, I’m a big fan of the microenvironment, had a lot to say about that for many years. But I guess this is a question for you, Becky, is, you know, every other cancer, next-generation sequencing, there’s a panel you should be running, you know, and whether you’re a lung cancer expert or a whatever cancer expert, you’re doing NGS across the board. What do you see as the value of the NGS panels in the world of large cell lymphoma? And then to the point that Paolo made, where do you think we’re going with trying to look at what else is around a tumor in the microenvironment and what should we potentially in the future pay attention to? 

Rebecca King:

Yeah, so I think that’s a great question and a great point. I think, you know, for some of those cancers, the value of the NGS is to identify very specific targets, you know, like EGFR that are targetable. Unfortunately, DLBCL so far has not become a sort of one gene disease. And it hasn’t even broken down in the way that acute leukemia, if you will, has, where there’s multiple sort of one gene or one translocation or one pathway disease. So it really seems to be, for the most part, a sort of multigenic and extremely complicated, you know, milieu of genetic aberrations, epigenetic aberrations, and, as you said, you know, microenvironmental factors that all play into kind of the biology and then ultimately the clinical presentation. So I think that’s the sort of basic explanation for why we aren’t routinely necessarily doing NGS in DLBCL. I think the value of, you know, doing NGS in DLBCL right now is primarily to identify those biologic clusters we’ve talked about and to be able to feed that data into one of these classifiers. And so we’ve already touched on, you know, do we really need that information, etc. And then a large issue, again, as I said, is that, you know, these classifiers were developed with, you know, not with targeted gene panels. And so it’s, you know, how big does your panel have to be, you know, we have a 45 gene, it’s a pretty tight panel for B-cell lymphomas here, whereas, you know, some institutions may have a 400 gene panel. And so, you know, how do those things play in? How much information do you need? And so, again, I think it comes back to the clinical demand. You know, we do run lymphoma NGS here and, you know, as a reference laboratory, we see it not only for our internal patients, but externally. And I can tell you that we are getting requests for NGS on DLBCL, but my experience is that internally, we’re not really getting requests from our hematologists on DLBCL. And, you know, just very briefly, diagnostically, although NGS absolutely has value in other lymphomas, it doesn’t really have value diagnostically in DLBCL. So I think that’s kind of where we are with it. 

Stephen Ansell:

Yeah, I must say again, I think such an interesting kind of moving target if you’re like. Paolo, I want to switch a little bit. You know, resampling the lymph node over and over and again is actually kind of challenging. And quite frankly, I’m sure Becky will be able to comment on this if she wishes. You know, the sample gets smaller and smaller and smaller. So the more we’re trying to actually get our heads around what the microenvironment looks like and whatever is harder and harder. So I think I’m seeing much more focus on things you can learn from the peripheral blood. So again, cell-free DNA, which is going to be an amalgamation of all these cells, more and more kind of being learned from the peripheral blood. So just give me your thoughts on what do you think cell-free DNA is doing for us right now or could do for us, both from the perspective of right at the beginning, tracking response to therapy, and then once you’ve had a response, the whole MRD kind of question. What do you think? 

Paolo Strati:

Yeah, I’m actually very optimistic about the potential clinical implications in the not so far future of circulating tumoral DNA-based type of assays, and I’m going to summarize what we currently do and what we could do in the future. To your point, I agree. Well, on one hand, there is a clinical need to try to resample every time a patient relapses to understand what targets are expressed for therapeutic selection. As you mentioned, it can be an issue for the patient for invasiveness per se. It can be an issue of sample quality. But also one thing that we didn’t mention is that in lymphoma, probably more than solid tumors, there is a lot of geographical heterogeneity. So that as we place our needle in a specific lymph node, we may not capture the global biology of the lymphoma, and so we may select the wrong treatment. So one of the advantage, one of the many advantages of circulating tumor DNA is that by sampling instead plasma, we may probably be looking at a more global and deeper biology of the lymphoma, and we may be able to do our treatment selection more properly. Now, with that said, what are we currently utilizing? Just to summarize, historically, what has been utilized for minimum residual disease evaluation of lymphoma has been initially very suboptimal techniques that may still be relevant in simpler lymphoma, like, I shouldn’t say this, but, you know, in circulating lymphomas, like mantle cell lymphoma or CLL, such as flow cytometry or qPCR. Then next-generation sequencing started, initially clonoSEQ, CAPP-Seq, and then what everybody keeps talking today about is, of course, PhasED-Seq. The latter has shown a lot of promise in the frontline setting. There are very robust data showing that patients with large B-cell lymphoma who receive frontline chemo if MRD-negative by PhasED-Seq at the end of treatment, have a very high chance to cured and not relapse five years later. Whereas almost invariably, those who were MRD-positive, again, by PhasED-Seq, specifically at the end of chemo, are more likely to relapse subsequently. And those are patients on whom we should probably try to focus for consolidation strategies. And there are ongoing trials like the ALPHA3 trial where patients who receive frontline antracycline-based chemo, they have a complete metabolic response by PET, but they are MRD positive by PhasED-Seq, they’ll be randomized to being observed or receiving novel treatments, in this case, allogeneic CAR. So this is already happening in clinical trials, and there are many trials that are currently being built or activated with a similar study design. Now, one thing that we have learned is that the regimen that we utilize may matter in terms of reliability of the MRD assay. So we keep being so focused on the sensitivity of the assay, and PhasED-Seq can identify one tumoral molecule out of a million as compared to clonoSEQ or CAPP-Seq that may be limited to 1,000, 100,000 molecules. However, sensitivity is not everything. So what we have learned applying PhasED-Seq to pivotal trials of, let’s say, CAR-T, so more dynamic treatments that may eradicate MRD over time, it gets weaker in that case. And let’s say, if you look into a patient with large B-cell lymphoma, gets CAR, and at a very specific point in time is MRD positive by PhasED-Seq, they may become MRD negative subsequently. So the treatment type affects the utilization and selection of the assay. 

Stephen Ansell:

Very helpful. So Becky, as we’re kind of winding things down here, you know, looking to the future. So you mentioned how, you know, using things in diagnostic arenas, you know, we’ve now got these methylation tests, these kind of multi-cancer detection things. And I’m starting to see these in my clinic, the executive health programs and those are kind of doing those. 50 different cancers, and then I get a little message that says cancer signature detected, lymphoid malignancy likely, or something like that. What do you make of that? And how do you see that being helpful in diagnosis or maybe it’s not helpful at all? What do you think? 

Rebecca King:

Yeah. So, I mean, my disclaimer is that I have very little experience with these or, you know, knowledge about them. My anecdotal experience is exactly what yours is, which is that I have seen a handful of biopsies come through in patients who had this test done and it said you, you know, have a high chance of having a lymphoid malignancy. And they went in and, you know, lo and behold, they do. Again, that is, you know, my N of like two or three examples. My understanding is that the sensitivity of these is not that high, so probably not something that’s going to be, anytime soon, at least detecting occult disease or, you know, everyone should get this test so that they screen for, you know, occult T-cell lymphoma or B-cell lymphoma, but it is certainly good at detecting, you know, when you have a fair amount of disease. I think, you know, I see it probably as less exciting in lymphoma than in many other solid organ cancers, because, unlike with many tumors, for the most part, diagnosing lymphoma a little bit earlier is not really markedly changing the course of that patient’s life, and in fact, may in some cases be detrimental. Whereas, you know, detecting pancreatic cancer earlier is a big deal and is something that, you know, a lot of people are interested in. So my bet, if I were putting money down, is that this is not going to take off in lymphoma anytime soon. 

Stephen Ansell:

Oh, that’s very valuable. And Paolo, your sense of, do you see any of these in your practice? And if you do, what are you doing with it? 

Paolo Strati:

Yeah, so I do. And actually, I completely agree with what Becky said, because I was going to say exactly the same thing. So my disclaimer is the same. I don’t request those assays yet, and my exposure has been for patients who did this with an outside provider and that prompted the PET scan and then biopsy and they come to me with an initial diagnosis. And I was going to make exactly the same point that Becky made; I can see how this may be relevant in solid tumors, where the kinetics of growth can be very slow, but then still a diagnosis of solid tumor early may be very impactful for the patient, as you mentioned, gastric cancer, pancreatic cancer. Now, a lymphoma, typically aggressive B-cell lymphoma, has such a fast kinetic of growth that it’s unlikely to remain asymptomatic for several months. So most of patients do have palpable lymphadenopathy or constitutional symptoms or compressive lymphoma-related symptoms way before seeking a test like this in absence of symptoms. And when it comes to slow-growing lymphoma, that is not the focus of today’s discussion, like follicular lymphoma, marginal zone lymphoma. As Becky said, many times knowing so early and in absence of symptoms that you do have marginal zone lymphoma, follicular lymphoma, with a very low tumor burden and no indication for treatment becomes sometimes just an unnecessary source of anxiety and also triggers unnecessary testing at that stage that could have started much later. So I don’t necessarily support its use, based on the current data and my very limited personal exposure, in lymphoma, but I can see the utilization in solid tumors. Maybe one thing that I may add is that one area of interest could still be within lymphoma for patients where based on the location of the lymphoma or the information that we can gather from the lymph node biopsy, we struggle to characterize the lymphoma. It’s not a degree of information that, to my best knowledge, this assay offers yet, but if you were able to characterize the lymphoma subtype utilizing blood, again, it may be helpful. And I’m thinking of situations like, let’s say, patients with large B-cell lymphoma that sometimes relapse with isolated brain lesions where we keep doing lumbar punctures and we don’t see anything by flow cytometry or cytology and where a biopsy is not feasible. It would be interesting to maybe utilize this assay for those patients where we cannot do a diagnosis utilizing a regular biopsy. 

Stephen Ansell:

Very helpful. So I’m hoping that those that listened to this discussion around modern diagnostics in diffuse large B-cell lymphoma have found the comments of both Becky and Paolo very helpful. Becky King, thank you very much for your excellent hematopathology insights. Paolo Strati, thank you very much for your fabulous translational and clinical input. Thanks to the audience who listened. I trust this was valuable. My name, as I said, is Steve Ansell, and thank you very much for participating in this meeting. 

Rebecca King & Paolo Strati:

Thank you.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.

Paolo Strati:

Research support: Sobi, Astrazeneca Acerta, ALX Oncology, ADC Therapeutics, Kite-Gilead; Advisory Board/Consultancy: Kite-Gilead, Novartis, Sobi, ADC Therapeutics, Incyte, Astrazeneca Acerta, Roche-Genentech, Abbvie-Genmab, Merck, BeOne, Lilly; Active grants: Leukemia Lymphoma Society (now Blood Cancer United).