Early screening in lymphoma: current perspectives & future outlooks

Jessica Okosun: Hello, I’m Jessica Okosun, and here we are in Lisbon for the Intercepting Blood Cancer Program 2024. And I’m from the Barts Cancer Institute in London. I’ve got with me Sandrine Roulland, who, Sandrine maybe you want to introduce yourself?

Sandrine Roulland: So I’m a research director at French NIH INSERM, and I’m working on follicular lymphoma and precursors.

Jessica Okosun: So, Sandrine, we’re in this really interesting meeting where we’re focused on whether there’s the potential to screen patients for blood cancers. And obviously both of us work in lymphoma. So I guess the first question I could ask you is, do you think we’re at that stage where we can screen patients or healthy individuals for their risk of developing lymphoma?

Sandrine Roulland: So I think my response will be different from two years ago. From two years ago, we were just looking at a particular translocation, the genetic hallmark of follicular lymphoma that is present in almost everybody, almost 70% of healthy individuals. In that case, it was not really, I think I don’t know, was not really interesting to propose a screening because too much individuals. But, I think now that we have deeply characterized the genomic landscape of those precursors, particularly we have used a very sensitive technique allowing us to tackle the very rare mutation present in few precursor cells. We are able to distinguish people that will develop follicular lymphoma that have more abnormalities than t(14;18), they have epigenetic abnormalities such as CREBBP or KMT2D. And the ones that got both mutation or, at least both or three mutations, including the epigenetics, they are more prone to develop follicular lymphoma, while in the ones that do not develop, they have just t(14;18). So we can think here to have a test that can eventually measure the frequency of t(14;18), which is a good indicator of progression and also eventually detect those CREBBP mutation. But the difficulty here, I would say for screening is that we need very sensitive approaches. We use ctDNA-like profiling with a sensitivity of 1 to 100,000 cells, which is very low and which is very expensive. So it can be a bit difficult to apply that to a large population.

Jessica Okosun: So I guess the question, Sandrine, is that, you know, one of the challenges with screening test is that we could potentially be identifying individuals who don’t develop lymphoma later on. So that could create a sense of maybe anxiety in the patient. So maybe we could be over or under diagnosing individuals. So how do you think, you mentioned there already that sensitivity is really important in terms of the test, but is there anything else that you think we could be doing to improve that?

To improve that? So I think we have listened to talks about the genetic risk, so the germline genetic risk. I think if we combine some genetic background, germline background plus the somatic mutation, we can have maybe something more ready to be applied to people, which consider both the somatic and also the germline risk. But it’s just an idea. I agree with you, I think I would maybe, I would prefer maybe screening in people in which we detect some hyperplasia. And we have seen that in some in situ follicular neoplasia that are really histological lesion, some of them will develop follicular lymphoma, but we really don’t know. But they are at risk because they have hyperplasia and permanent hyperplasia. And in those ones..

So focusing on high-risk focusing population.

Yes, focusing on a high-risk population.

So I think it’s because lymphoma is probably a little bit different from some of the other blood cancers. You know, when we think about multiple myeloma, the precursor being MGUS. And then we think about some of the leukemias, the precursor being clonal hematopoiesis. The kind of evolutionary trajectory is a bit clearer in those precursor conditions. Whereas I feel that in lymphoma we’re still probably a little bit further away in the sense that it’s not, the risk of transforming from this sort of precursor state to overt disease is probably not quite as clear cut as the other.

I think even if you have the, as I say, the two mutation, (14;18) and epigenetic, it could take ten years to get the disease. So I think we are we are really far from MGUS, where there is a certain population in the blood that is detectable easily, I would say detectable. But in our case it’s not yet there. So I think we are kind of yes, maybe a little bit behind. Yes. I think what we would need to do now is, we are talking about genetics and we know that genetics is not alone. We need to consider the immune microenvironment or at least the immune component of the individuals and this is not known for the early stages because we don’t have access to samples. And also we need to know the states. So maybe understand better the heterogeneity of those precursors, in terms of transcriptomic landscape, in terms of epigenetics landscape. And we don’t have the information. So we need to work in that direction. Yes.

So if I move the conversation a little bit from, from early diagnosis to, in lymphomas, one of the things that we can see is that you can treat a population of patients with exactly the same lymphoma, and some of them can stay in remission for many years, and others have a higher risk of either their disease relapsing quite early or transforming. Do you think that that may be a better time point for us to try and predict, because these are, you know, to try and identify these high-risk individuals and just want to hear your thoughts about whether you think that might be a better time point?

I think the diagnosis is still a good time point, but we have only access to one lymph node, and we know that there is a special heterogeneity in the follicular lymphoma. So maybe what we are looking at the cells that we derived to the relapse is not in the lymph nodes that we are looking at. And so we should have a more global view of the lymphoma heterogeneity at diagnosis with both bone marrow, lymph node, and certainly blood, to have a better view of what are really the precursors of the potential relapse and have maybe better biomarkers. So I think we need to work on other compartments than lymph node. They are the ones that are easily accessible. But I think we need to have a better view. And also we have some opportunities, I think, with people that are followed over time. So ctDNA profiling is actually looking at the genetics of the minimal residual disease. But we are not really looking at the cells from a transcriptomics functional point of view at residual disease. And I think there is here an opportunity when people are in remission, so access to sample is difficult, but in remission potentially there is some few very rare cells that can make the seed for the future relapse. And I think there is another precursors here that we can look at, and that will be, I think, the focus of the of the next studies.

Okay, so if I were to summarize Sandrine, so are you in the yes camp or no camp for are we ready to screen at diagnosis or healthy individuals to predict whether let’s say..

I would say not really ready.

Yes, I would say the same also but there’s probably more potential for maybe the later stages, as you say, maybe using after the disease..

Just after the first treatment, if there is something left, eventually, if we characterize better, we can come back with a new therapy at that stage when people have not yet relapsed. But that’s also a guess. But right.

Thank you Sandrine, very interesting to hear your insights and your thoughts.

Thank you.

AI-REAL: improving diagnosis and treatment in EBV-associated Burkitt lymphoma

Anna Schuh: I’m Anna Schuh, I’m a professor of molecular diagnostics at the University of Oxford, and I also spend some of my time mentoring hematologists and clinicians in Uganda and Tanzania.

I was very fortunate to receive funding from the UK government a few years ago to run a program called AI-REAL; the Aggressive Infection Related Lymphoma Study in East Africa. And as part of this study, we had three objectives: number one was to increase the speed and accuracy of diagnosis of children diagnosed with EBV positive lymphomas, in particular Burkitt’s lymphoma. The second objective was to evaluate whether the addition of rituximab to the chemotherapy would improve outcomes. And then thirdly, we were interested in, doing whole-genome sequencing of the Epstein-Barr virus strains that are isolated from those children with Burkitt’s lymphoma to see whether we could use this information for rational future vaccine design.

So what we realized at the beginning of the study was that we had to use a before after design. So basically, initially documenting really carefully what the current state of the art and standard of care, diagnosis and treatment were and what the outcomes were. And then in the after study, we would then introduce, a better diagnostics and also, provide children with rituximab. So when it comes to the before study, we very quickly realized that there were huge delays in diagnosing these children, and this is primarily because of healthcare related delays, both in primary care and in secondary care, although in secondary care, most clinicians are aware that Burkitt’s lymphoma is an emergency and needs a very fast treatment. The problem was with the provision of pathology, there’s a huge lack of trained pathologists and also a huge lack of immunohistochemistry and other specialist technologies that we need for the diagnosis of Burkitt’s lymphoma.

So in order to address this, we developed a liquid biopsy test essentially looking for circulating tumor DNA and circulating viral DNA in blood samples from children.
And we developed this test and then introduce it in the diagnostic pathway, and we managed with a median time to diagnosis of 11 days. The fastest diagnosis we established was six days from the northern Uganda, the border with the South Sudan and the Congo, all the way to Kampala in six days. That included the sample transport, processing of the sample and issuing a diagnostic report via the multidisciplinary team meeting that we run virtually. So this was a very nice success story.

We then looked at giving rituximab to the children and essentially what this showed was that rituximab does indeed improve event-free and even overall survival at one year of follow up. The rituximab was well tolerated, and it’s feasible. We’ve done health economic analysis, some of this is still ongoing, but we think that both the liquid biopsy and rituximab are actually cost effective because we are dealing with children and the number of lives saved is very much increased.
We also think that especially on the diagnostic side, we can very much improve on the cost by scaling and also by using, probably different, more diverse sequencing providers.

So, when it comes to the vaccination story, we are currently in discussion with the Oxford Vaccine Group to use the ChAdOx2 version of the Covid vaccine that was developed by Oxford, to actually vaccinate against Epstein-Barr virus. So in this effort, we will try to vaccinate children within a clinical study after they’ve achieved complete response to chemoimmunotherapy and then provide four vaccinations over a period of one year as a therapeutic anti-EBV vaccine. With respect to the liquid biopsy testing, all the analyses were obviously done both in Tanzania and Uganda.
We’ve created a social enterprise called ries are called SEREN in Dar es Salaam and this is essentially a laboratory that specializes in cancer and blood disease diagnostics. So we’re very keen to also evaluate the use of multi-cancer detection testing alongside the more specific cancer testing that we’ve developed for Burkitt’s lymphoma