Aleksandra Zuraw: [00:01:34] The quest to cure cancer is ongoing and currently with cancer immunotherapy, also known as immuno- oncology scientist, are tapping into the power of our own immune system to defeat cancer.
[00:01:47] If you were ever wondering what cell populations and processes play an important role in this approach and how image analysis helps the researchers, this episode is for you.
[00:01:58] Today my guest is Professor Elfriede Noessner from Helmholz research center in Munich, Germany. She is a member of Visipharm’s scientific advisory board and an immunologist with over 35 years of experience in the field of cancer immunotherapy and she will give us an overview of the interactions between the tumor and the immune system and the role of tissue image analysis in the process.
[00:02:25] Hi Elfriede, how are you today?
[00:02:27] Elfriede Noessner: Thank you. Very well. Yes, I’m looking forward to our talk.
[00:02:31] Aleksandra Yes. Thank you for joining me on the podcast. Let’s start with you. Let’s talk about you. You’re an expert in immunology. Tell us a little bit more about your background and about your research, and how you got to that place.
[00:02:47] Yeah, I guess I’m an expert in immunology. I have done that since 1985 when I did my PhD work in the lab of Professor Schendel, and we studied the T-cell response towards antigens at a time where there was not so much understood about how that actually works. I continued doing immunology research on antigen presentation and T-cell response during my postdoctoral thesis, which I did at Stanford University.
[00:03:20] There, I studied the immune response towards transplantation and my former employer said, I shouldn’t switch the topic. I should stay with transplantation because tumor immunology has never worked. Times have changed, and it’s now about the most upcoming therapeutic approach in cancer immunotherapy and also in autoimmunity, where it was actually always an accepted discipline. My working place currently is the Helmholtz Center in Munich.
[00:03:52] My main topic or my research field is the renal cell carcinoma, but also lung cancer, and it’s trying to understand how the immune response works or actually is inhibited against the tumor in order to make it happen again after it has stopped in the tumor microenvironment. Yes, that’s my background.
[00:04:16] Aleksandra: So, why did you pursue this field of immuno-oncology, even though your supervisors or your co-researchers discouraged you from that? Why did you think it was important then, and why is it important now?
[00:04:37] Elfriede: I guess, easy answer is I was always stubborn, and I was always interested in difficult questions. At the time when I joined, the tumor immunology was actually gaining a bit of ground because it was about the time when the first two antigens in human tissue actually were identified. It was the group by Terry Boon in the Netherlands. So, the tumor-immunology aspect of the topic gained a bit more scientific and got away from the anecdotal field, which it was called.
[00:05:15] It’s anecdotal that in your responses towards tumors happened sometime, but nobody understood why. So, this was interesting, and we had a very good connection to the urologists, and it was well-known that renal-cell concept can disappear spontaneously.
[00:05:36] So, it was thought that this can only be the immune system, because what else can make things disappear. We do know that the immune cells, they can destroy things. Pathogens, it’s very well-equipped to do that. So, it was indications that the immune system can be directed towards tumor tissue and that kind of interested me.
[00:06:04] Aleksandra: So you say, obviously, immune cells can destroy pathogens, and you suspected back then already that the disappearance of renal cell carcinoma can be due to immune system. What are the immunological processes and also cell populations?
[00:06:22] You said you were working with T-cells at the beginning of your career. What are the processes and cell populations relevant for oncology, for immuno-oncology?
[00:06:33] Elfriede: Yes, the processes are, of course, in the tumor field is killing, removing the tumor. I mean, if you think about auto-immunity, it’s actually the opposite. That was also very interesting to me that this auto-immunity in tumor immunology may be two sides of the same coin. So, understanding the principles, I thought, and I think that is also true, will help to understand better any of the disciplines.
[00:07:04] Auto-immunity will help the tumor immunology, and tumor immunology will help the auto-immunity. Because, in the one side you want to enhance the immune response, and the other side, you want to stop it. So, you can learn from each other. The processes, I mean, the immune system is very complex. It has different players, and they all do something different, and they work together.
[00:07:34] So, the killing process is probably the most relevant for the tumor immunology, but it’s equally important that you remove the debris, because if you get an accumulation of debris, then the organ which you actually want to resolve, or you want to heal is going to be destroyed with the debris.
[00:07:58] So, an important, equally important process to killing is also removal of debris. Here, another cell type is important, like the macrophages, you do know that those certainly these are our eating cells and so need to have macrophages as well. Other processes you do know, are antigen presentation, but does that mean it is you have to enable T-cells to see the enemy, basically.
[00:08:36] So, you have to have a certain cell type which presents and shows the antigen to the T-cell so that it can launch an attack. If you do not have antigen presentation then the T-cell is blind, it does not know what to attack. So, the antigen presentation is done by a specialist subtype, which is all the dendritic cells.
[00:09:02] So, now we have T-cells, we have the dendritic cells, we have the macrophages, then there are the B-cells, antibodies and the B-cells produce the antibodies and the antibodies bind to the antigen, that’s why it’s an antibody. So, by that, it marks the cell that should be destroyed, for example, by a macrophage. So, B-cells are also an important part. The B-cells in the tumor field, they are upcoming as an interesting cell type.
[00:09:38] The processes are not quite understood what it will do with the tumor tissue, because it’s a systemic response, and the tumor, at least for the most part or solid, they are organ localized on the antibodies, they are in the situation. So, the B-cells that is sort of developing, but has not necessarily defined mechanism behind it in the tumor field.
[00:10:14] So, we had the killing, we had the eating of the debris, antigen presentation, and then equally important is the regulation of the whole communication and that stopping the immune response. This is a very important process in the immune system because it prevents overshooting and wrongly reacting immune response, which leads to auto-immunity.
[00:10:42] The tumor field is regulation, which is a built-in process in the immune response. This regulation is coming too early. It’s stopping the T-cells before they have removed all the tumor cells. So, the regulation, the stopping process, which is you cannot live without the stopping reaction in the immune response, it’s detrimental in the tumor immunology field.
[00:11:08] So, that were great breakthroughs to identify those mechanisms that stop the immune system. It’s like, you do know checkpoints? They call checkpoints that point, it’s controlled, does the immune response go on so does the killing proceed or is it stopped? These checkpoints are a CTLA-4 and PD-1/PD-L1, which have led to these advanced therapies, which are now applied to a lot of patients and where there’s great benefit in the tumor control by stopping or overriding those checkpoints.
[00:11:51] Especially, PD-L1 is something that is really, popular is maybe not a good word, but commonly present in the immuno-oncology field and also something that you can detect in the tissue.
[00:12:01] Aleksandra: So, my next question is what role does the visualization and quantification of those cells and those checkpoints in the tissue play? What is the role it plays in advancing immuno-oncology or immunology? Does it merit? How relevant is it to see it in the tissue?
[00:12:23] Elfriede: One easy answer is, seeing is believing. You can convince somebody right away if you can make them see. So, that’s a simple answer, but of course, this is also can be documented by more scientific ways like you see that the T-cell is in the tumor tissue, then you do know, at least has found a place or the right localization where it should execute its job.
[00:12:56] So, it has to be on the right location, and you can visualize that when you do an image analysis and the great breakthrough also for tumor immunology are that the immune cells, the T-cells actually can make a benefit or to have a benefit in tumor control was a paper in Science.
[00:13:19] So, the magazine Science published in 2006 by Galon Jerome, he was able to visualize the T-cells in the colorectal cancer tumor tissue, and he could demonstrate that by counting the number of T-cells that matters to the prognosis of the patient. So, what he was able to show is that if there are many T-cells in the cancer tissue then the patient has a better prognosis.
[00:13:53] He did that in a patient cohort, and he was able to show that the prediction of prognoses was better based on the T-cell numbers than it was on the pathology like it was done before, or it’s still done, but he could show that in the situation with the stage-1 tumor, for example, where this is a very good prognosis in general, but there are some patients still in stage-1 who relapse early, and he could show that those patients who do not relapse, really do not relapse, they have a high count of T-cells in the tumor tissue and those which relapse, even though, they are stage-1, they have a low count.
[00:14:42] Aleksandra: I remember that work. This led to this Immunoscore, it’s kind of a test now. Right? Image analysis was used for this test. But, go ahead. Go ahead. Keep explaining because this is fascinating.
[00:14:55] Elfriede: Yeah. So, this was sort of really 2006, making even pathologists believe that T-cells matter, can matter. I have to say because they do not work all the time and that is something where this is currently struggling, or that must be the future to understand why they do not work all the time.
[00:15:23] Here, I come back to the regulation. The immune system stops itself after a while. This is absolutely required, otherwise, tissue will be destroyed that should not be destroyed. So, to understand what regulated there, are different levels. One is the checkpoints PD-1/PD-L1. Then, you have also regulatory cells, the regulatory T-cells, for example, and those can be also in the tissue, or they are most of the time.
[00:15:58] Here, it’s also important not only that they are in the tissue, but to regulate the CD8 killer cells, they must be in close proximity to the CD8. So, if you have the same number of TREX1 tissue, but they are far apart from the CD8 T-cells, and that doesn’t matter, it doesn’t inhibit the T-cell. If they are close together, they inhibit the T-cell. So, you have two different situations, same number, same type of cell.
[00:16:31] The one sit together, the other ones sit not together. You have a different outcome. In one situation, the CD8 cells are inhibited, in the other not. That you can only determine when you have an image, because if you dissociate your tumor tissue, you can quantify precisely, and you can define which cell types are there, but the Proximity analysis, you can never get with any other analysis than an image analysis.
[00:17:02] The outcome as I explained, it’s very different whether the cells sit together, talk to each other, or they do not sit together. So, this is one example where image analysis is the only way to determine this information and that has consequences for understanding why the immune cell, the CD8 that’s not killed or why it does killing, another situation, for example.
[00:17:33] So, the visualization of those cells, obviously, we have immunohistochemistry, immunofluorescence, and multiplex immunofluorescence. Are there any other methods that you are using in your research?
[00:17:48] At the moment, these are sort of the ones becoming established. The single-marker analysis, I mean, it’s done by histology for, I would say, centuries. So, that’s the most simple and the easiest one that gives your part of the information, because you can always determine one cell type or one process, one factor. If you have a multiplex standing, this offers much more information from the tissue.
[00:18:23] You can see the different cell types in one section slide and this gives you the information for the communication. As I said, that immune system is a system that duplicates, and it matters what they say to each other because they may say, “Stop” or they may say, “Go. Go faster.” So, if you have multiple markers that which you can stay on the same slide, you can get much more of this information.
[00:18:58] It’s however challenging because you have to be able to really, precisely distinguish the cell types, you have to have good markers, you have to be able to sensitively detect them, and the anions have also to make the analysis. The more markers you combine, then your visual brain that’s no longer able to kind of make all these connections, how many cells are there, how many of those, which are there or in contact, and so on.
[00:19:33] Each marker you add makes it more complex for the analysis. Then, you cannot do it by eye anymore because you just cannot grasp the complexity. That’s one, you have to get into a computer-assisted detection system and have to actually trust a non-human being for yourself.
[00:20:01] Aleksandra: So, is it a common practice now all when you have multiple markers image analysis used period, right? There’s no attempts or do people still try to just do it by eye or assess semi-quantitative way without using image analysis? What’s your experience in that?
[00:20:24] Elfriede: I can say for myself, I want to always see. So, I’m the one who kind of wants to see and does not want to go blind into numbers.
[00:20:35] Aleksandra: No. The same for me, I have to see this, because I want to know that the algorithm is doing a good job.
[00:20:43] Elfriede: Yeah. I’ve worked with cells, I worked in cell culture and I like to see things, but you have to get beyond that. You have to be able to kind of charge the machines at, I say this way, so that the algorithms are correct, but you still have to…
[00:21:11] There’s a step where you first control to see and supervise so to speak, but there is no other way than to go beyond the visual analysis, that there is no way that to sort of avoid that or to not do that because otherwise, you can not really grasp that the different communication processes that are present in the tissue and the different cell types, how they react with each other because the system is complex, and you cannot ignore that complexity.
[00:21:57] Aleksandra: So, another question that I have… Let’s say we have a biopsy or a resection of a tumor. This is a section of a point in time, obviously, immune system is super dynamic. How reliable is this information if we only have it from one time point?
[00:22:19] How many would we need to assess the dynamic? Is it even possible in the current immuno-oncology field? Maybe, research is making it possible, but would it be possible in the clinic? What’s your take on that?
[00:22:36] Elfriede: I mean, this is a very important question and there are lots of attempts to understand what is really required to know. So, multiple biopsies taken, how many people really grasp the heterogeneity of the tissue, which is present. So, there is no doubt that the tissue is very heterogenic.
[00:22:59] So, there are comparisons made on the molecular level and also on the cell types. So, for the immune response, I am not so extremely worried about the heterogeneity because the immune system is dynamic. I have to find a point where I can launch that immune to direct the immune system to attack and from then on, it may develop itself. It’s like starting in a little snowball and that it becomes an avalanche.
[00:23:42] So, I have to have one attack point. Let’s, as an example, the T-cell destroys this one cell where I directed to, and then this cell releases other factors, other antigens, and then new T-cells will attack these second antigens and attacking the second tumor cell, which may or may not have the first antigen at all. Then, the second tumor cell line will release another antigen A, B, C, and D and against those, the T-cell again will develop.
[00:24:23] New T-cells will develop. So, with directing the immune system towards the tumor target, initially, I do not have to have a broadly expressed target because these other targets, which are hidden in the tumor cells initially, they will become obvious to the immune system. But what I need to have for this immune system to develop is, as I introduced before, I need antigen presentation.
[00:24:59] So, these newly-released antigens with my initial T-cell wasn’t seeing those. They have to be made visible to the other T-cells, which can react them. So, it’s important that I have this antigen presentation ongoing during the immune response. Therefore, the dendritic cells may really be one of the central cells in this process that T-cell is the executor, but to have them ongoing, I need these presenting cells.
[00:25:44] Therefore, I guess one has to look for the presence of this antigen presentation cells in the tumor tissue, or also it can happen in the lymph node as well. So, the heterogeneity is not so critical, in my opinion, for an immune response towards the tumor. It’s very critical, of course, if I have a molecular target, a drug target because if it’s like and directing it puts the mutant K-Ras or so in lung cancer, then that my drug can only detect those cells, which have this one particular mutation and the drug cannot change, it cannot detect then another mutation that is in other cells.
[00:26:36] So, for these targeted therapies, this molecular targeted therapies, the heterogeneity is really critical because if I target a mutation that is present in only 1% of cells, then I kill only 1% of cells. There is no way that I can kill the other ones by the same mechanism. Therefore, heterogeneity is very important for targeted, molecular-targeted therapy, but not so critical for the immune targeted therapies.
[00:27:13] Aleksandra: I never look at this from that point, but you’re totally right. This is a dynamic system.
[00:27:20] Elfriede: What still happens is that the immunological environment will also change when the immune system is starting to do something. So, a T-cell when it attacks its tumor cell effectively, it not only kills, but at the same time it releases cytokines, the most important or most common one is Interferon gamma and those Interferon gamma does something to the tumor cells.
[00:27:53] For example, it induces PD-L1 and PD-L1 prevents the tumor cells from being killed by the T-cell. So, I have a process of self-inhabiting or self-limiting immune response, which is our built-in regulation. So, a T-cell kills its target produces Interferon gamma, so the neighboring tumor cells will be PD-L1 positive and inhibit the incoming T-cell response.
[00:28:24] So, I have to foresee, basically that was an effective immune response, I get a very responding inhibition. So, in my initial analysis of the tumor tissue, before the T-cell attack happens, the PD-L1 may not be even present, so I say, “I have no PD-L1 inhibition.” But that’s not true. It may occur after a half a certain immune response initiated it.
[00:28:56] So, I have to foresee what will happen if an immune response occurs. So, what I do know from immunology is that if I induce a T-cell response that I can count on that it will have subsequent inhibition, so this inhibition I have to target as well so that the immune response can go on for a longer time. So, this dynamics I kind of have to foresee, and I can only foresee when I understand what processes I initiate, what’s my initial therapy. I thought it’s all, it’s not so easy, and it’s not static.
[00:29:50] But, basically the one image… Let’s say, we take one image and this one image gives you a starting point from which you start interpreting and thinking, okay, this is the therapy I want to use, this is the processes I want to start, then subsequent processes are this and this and that and because it’s a dynamic system, it’s not crucial though, have biopsies every other day or something like that.
[00:30:21] I would say not however it would help if one would have a biopsy later on, because after all, it’s still a hypothesis, what I foresee, although I say I’m pretty sure this will happen. Pretty sure is not enough. After all, we have to know.
[00:30:45] So, in the process or at the point where we are at the moment, we do not have that data whether what I foresee is really happening and in that, for that reason, studies are required for sequential biopsies are taken and are analyzed and that will provide us with more information how the immune system actually develops and where we have to intervene further on, that we may be able to actually build a model, a mathematical model where we can foresee what will follow after one and another, how the waves will occur, but we do need the data for that.
[00:31:35] So, in that regard, sequential biopsies or study-support sequential biopsies are taken are actually absolutely mandatory to set up those studies, to approve them and execute them because that the information is scars on that. It’s just my interpretation from how immune cells and immune the immune system regulates itself. But the data, particularly in the human system, are still not very commonly done. It’s all from mouse work, and we can predict things, but we again have to see whether that really is correct what we predict.
[00:32:30] Aleksandra: Definitely. I did, I worked in the immuno-oncology image analysis space. So, I’m not totally new to the subject, but this has been a great overview for me. I have benefited a lot from talking to you, Elfriede. So, thank you so much for making this pretty complex topic clear to us and to everyone who’s listening.
[00:32:55] Elfriede: It was my pleasure. I’m fascinated by the immune system. I’d like to communicate this fascination and people are afraid that this in unity is so complex, it is complex, but there are logics behind it. We have to understand that we have to observe, and we have to study it.
[00:33:15] Image analysis, I mean the new technologies, they offer great opportunity to better understand what actually goes on in the tissue itself. So, in that way, I’m very thankful that people develop those analysis systems, and I’m very excited to apply them and to see how far we can take them and how we can improve.
[00:33:44] Aleksandra: Definitely, something that has been helping immuno-oncology for quite some time already. Thank you so much for your time and for this great conversation. I hope you have a great day.
[00:33:56] Elfriede: Yeah. Thank You very much for your time.
