iwNHL 2023 | Updates on the biology and treatment of T-cell lymphoma

Francine Foss: Hello, I’m Francine Foss, and I’d like to introduce you to my colleagues here at the iwNHL at our T-cell session, Francois LeMonnier, my co-chair, Laurence de Leval, and Dr Steve Horowitz. So I would like to ask my colleagues here to summarize briefly what they thought the important points were from their talk. You talked a little bit about the background of mutations and T-cell lymphoma, and what’s going to be important to us as clinicians. Could you tell us a little bit about that?

Laurence de Leval: Yes, of course, Francine, so I provided some overview of the entire accumulated past ten years and genetic sub-stratification and characterization of lymphomas. I talked a little bit about the historical group, that of anaplastic large cell lymphomas ALK positive versus ALK negative, that distinction still matters. But, the question now is: what is the significance of the genetic variants of ALK negative ALCL? And the situation is not so clear on the relevance of the testing for DUSP22, while rearrangement, while TP63 rearranged cases are generally aggressive, but they’re very rare. They seem to be other variants like JAK2 rearranged which are less well characterized, so there’s still a way to go. In addition, those cases have frequent mutations in the JAK-STAT pathway. That’s another way of looking at the mutations and abnormalities- by pathway rather than by entities. And among the signaling pathways that are important in pathogenesis with potential therapeutic intervention are the JAK-STAT signaling pathway often activated by mutations or rearrangements, and the T-cell receptor NF-κB pathway, also supposedly mostly activated by rearrangements. These abnormalities you see in various PTCL entities. That is the background for constructing clinical trials that are entity agnostic but based on the molecular alteration. And I think the JAK-STAT pathway has the best characterized inhibitors. So that was developed further by my colleagues.

Francine Foss: Yes- Steve- you spoke some about JAK inhibitors. You gave us some clinical data from some of your studies.

Steven Horwitz: Yeah. So what I talked about today was really looking at a couple of signaling pathways JAK-STAT, PI3-kinase, ITK, and then some of the preliminary clinical data. So, in targeting the JAK-STAT pathway, we did a Phase II study of ruxolitinib, and we were able to see that there is some activity with targeting that pathway. Activity is most in patients that have alterations of that pathway, with mutations or elevated phospho-STAT signaling. So it looks like, by targeting that pathway, you can have responses, you enrich responses. A lot of work is looking at trying to correlate how much we can predict who’s likely to respond, who’s not going to respond, so we can individualize the strategy.

Francine Foss: Francois, you talked a little bit about the whole epigenetic landscape and how we could potentially target that.

Francois Lemonnier: Yes, indeed. Actually, we know that for more than ten years, I think, between 10 and 15 years, there are very recurrent mutations in T-cell lymphomas, especially in the [inaudible] T-cell lymphomas, with mutations in TET2 in more than 80% of patients which is an important epigenetic change, especially in the DNA methylation and hydroxymethylation. So it makes a rationale to develop epigenetic drug therapies, including with azacitidine. So we met with the French group, LYSA, but also with other collaborators in the UK and with the Nordic Lymphoma Group, in Japan, and we made a Phase III study that randomized the oral form of azacitidine (CC-486) in the experimental arm, randomized against the control arm with drugs at the choice of the investigator between romidepsin, gemcitabine or bendamustine. So, the trial showed interesting results with an increase in overall survival and PFS, progression-free survival, which was the primary endpoint, actually, was not reached, likely due to a lack of power of the study, because there was only 43 patients treated with azacitidine. We observed some sustained response because some patients are still on therapy now, four years after the initiation of the treatment. And, so, we want to try to understand better how to improve those results. The safety profile of the drug was really good actually. And so that is space for combination. We really want to investigate again this drug in combination in those lymphoma patients.

Francine Foss: I talked a little bit about valemetostat which is also is an epigenetic modifier of some sort. It looked like the response rates were very high with Valemetostat across different subtypes, which is something that we haven’t necessarily seen with some of our other therapies. And I know this touches on something that Steve talked about, which is that we’re now starting to dissect and look at specific subsets- rare subsets- and identifying pathways forward for some of those. And Steve, if you want to briefly just tell us a little bit about maybe some of these orphans that you have seen activity for in some of these new drug trials that you have been doing?

Steven Horwitz: Yeah, And I think the best example we have of that right now still is with targeting JAK-STAT. So, not us, but others, have identified that some of these tumors are really characterized by these frequent fusions, particularly in JAK2. And there’s a whole class of tumors that have this, including some rare cutaneous T-cell lymphomas, the aggressive epidermotropic. We’ve seen them in some PLL, other kinds of PTCL, and we see these sometimes in LGL. These rare or ultra rare subtypes that have these specific mutations, it looks like some of those patients, not all, are exquisitely sensitive to JAK inhibition. So, we talked about this idea that there may be molecular genetic subtypes where there may be a very effective, specific, on-target treatment. And I think the ruxolitinib data is sort of proof of that concept. But there’s probably other ways of identifying these and other ways of going about that. Then there’s a lot of discussion about the challenges in T-cell lymphoma, and how do we generate good enough data in these ultra-rare subtypes to generate either clinical use or payer support or even approval. And that is, of course, a big challenge.

Francine Foss: Right, yeah. And Laurence, if you want to summarize for us some of the data with animal models and how pertinent they are to us in the future, and also whether or not you feel we’re in the era where every patient with aggressive T-cell lymphoma should have deep sequencing done to identify these mutations.

Laurence de Leval: So it’s many questions in one Francine, I assume that you’re making reference to the talk that was provided by Dr Inghirami from Cornell. So, he presented us an update of data he has already partly explained a few years ago. As I understand, they have constructed a multi-dimensional platform using human samples and trying to create patient derived xenografts in mice, that was the basics. He has explained to us that by doing so, there is a high attrition rate with at least 30% attrition from what he explained. But they successfully obtained multiple PDX models of angioimmunoblastic T-cell lymphoma, for example, which is really the disease that is very common and that has no cell line, but also a lot of microenvironment that is retained in those mice. So that really provides a good experimental substrate to test for drugs. I’m taking the liberty because I’m sitting by Francois. I believe that he has a similar approach and has also successfully constructed some xenografts for drug testing in ex-vivo and can make additional testings. Giorgio Inghirami also explained that he could manipulate native human cells with specific mutations and create tumor models in humanized mice. He explained the complementarity of these different models in the platform. Now, to what extent this platform can really be used in real life to adapt therapy, that so far he has not really shown to us that this was done on a daily basis, or could be done in due time on a daily basis. Maybe you want to comment on that, Steve, because it’s close to you.

Steven Horowitz: Yeah. We’ve worked a lot with Giorgio. I think Giorgio has been fantastic at developing these models, and then we can test them. We haven’t done it real time concurrent with developing model, testing drugs, and then going back to the patient. What we’ve seen is, if you look at what happened to the patient in terms of sensitivity or not to therapy, you can often recapitulate that in the PDX, but not yet providing real time clinical information yet.

Laurence de Leval: And Francine, there was another part to your question. Do I believe that patients with T-cell lymphoma should undergo genetic testing? I would say I believe so, yes. I mean, it’s also my interest, but I do it at diagnosis. I think we learn from that approach, to learn about those patterns, to find out some associations. I think any single observations have value in this group of diseases that are very heterogeneous, the same way as Steve said, sometimes there’s very limited response, but it’s one element, it’s one data point. Similar in diagnosis, we see a lot of variability, and I think it matters. We accumulate data and then maybe find, relevant subgroups or limited numbers of individuals with similar patterns that may be relevant.

Steven Horwitz: If I can just add to that, I think the practical aspect that we see also is, in addition to learning more and more of our choices at relapse, they are not directed by, but informed by molecular testing. So if you get it at baseline, you have it when the patient doesn’t respond to therapy or relapses. If you wait till the relapse, you don’t have that information in real time. So we think there’s a practical benefit to getting a baseline, in addition to so much more understanding of the disease

Laurence de Leval:  And helping diagnoses.

Francois Lemonnier: Yes. I think it’s really helpful for the diagnosis because it really helps to have a better classification. And, I think we really made a huge progress in the understanding of the T-cell lymphoma pathogenesis. So I think now the molecular landscape is known. Probably, we need to learn a bit more regarding the interaction between the lymphoma neoplastic cells and the microenvironment. I think we still have a lot to learn. And now I think the emergency is to find some drugs, because now we have the mutations, we have a better classification, and a better understanding of the oncogenesis. So now, we need to find some good drugs.

Francine Foss: I think this was a really great session because we really brought the science and the clinical together and really talked about the translational elements here that are going to be important moving forward. And I’d like to thank all my panel members for their intelligent and really enlightening discussions here today, and look forward to our session next year. Thank you.