Hello, everyone, and welcome to today’s webinar, Development and Analytical Validation of Prognostic biomarkers for metastasis.

I’m Susie Valdez of LabRoots, and I will be your moderator for today’s event. Today’s educational web seminar is presented by LabRoots and brought to you by Visiopharm.

To learn more, visit them at Visiopharm.com. Now we encourage you to participate today by submitting any questions you might have during the presentation.

To do so, simply type them into the ask a question box and click send. We’ll answer as many questions as we have time for at the end of the presentation.

You may also submit any technical issues here as well if you have any trouble seeing or hearing this presentation.

I’d now like to welcome our speaker, doctor David Entenberg, an assistant professor at the Department of Anatomy and Structural Biology at Albert Einstein College of Medicine.

David, welcome. You may now begin your presentation.

Well, thank you for that introduction.

So as you heard, my name is David Entenberg, and I come from Einstein College of Medicine where I serve as the director of technological development within the Cross Lipper Biophotonic Center and its integrated imaging program. And today, I’m gonna be telling you a bit about the efforts that we have going on at Einstein to develop and analytically validate prognostic biomarkers that are specific for metastasis.

And so as a summary of what I’m gonna tell you, I’m gonna talk a little bit about the preclinical research that we have done over the years in order to discover two new biomarkers that are specific for the process of metastasis.

And these are then different from traditional biomarkers which really tend to measure proliferation and growth potential of tumors.

These biomarkers that we’ve discovered are really mechanistically linked to dissemination and metastasis. So I’m going to tell you a bit about how they were discovered and how we perform some clinical validation of these biomarkers and how we’ve been using digital pathology to automate the tests for, high throughput and eventual clinical use. How we’ve been showing, reproducibility and analytically validating these these new biomarkers. And then I’ll wrap it up by telling you, some of our latest work, which is how we’re using digital pathology to improve upon the prognostication ability of of these different, biomarkers.

And so to jump into it, I’ll first talk about how we’ve discovered these new biomarkers for metastasis.

So we’ve done, cancer research at Einstein for quite a number of years, and we’ve taken traditionally taken a pretty unique approach to cancer research. And that is we rely pretty heavily on intravital multiphoton microscopy.

We really like this technique because multiphoton microscopy has the ability to illuminate a single slice in an otherwise whole live tumor and create images that are very reminiscent of what you would get from mechanically sectioning and staining tissues, like in a in a standard histopathology lab. And that that technique can then reveal histological, features, but it also has the ability to, reveal the dynamics of cells since it is imaging inside of the live tumor. And what you see on the right here is a video that we we took, inside of a primary breast cancer, And what you can see is the tumor cells in green and the macrophages in blue.

And, the multiphoton microscopy, because it’s imaging live, can really give you the dynamics of of that, of those cells. And, by imaging many, many of these tumors, what we realized is that macrophages and tumors, tumor cells actually travel together inside of primary tumors, where they migrate in, directed streams toward blood vessels.

And when they arrive at that blood vessel, they tend to stop their motion and dock onto that blood vessel forming a structure that we call the tumor microenvironment of metastasis.

That is the green tumor cell, the blue macrophage, and the red endothelial cell that makes up the the blood vessel wall.

And these, we discovered by doing more intravital multiphosphate microscopy, we discovered that these, macrophages inside the TMEM structures can poke and prod at the blood vessel in order to open up the door, so to speak into the vasculature. And when that happens, you get this, leakage of vascular contents into the interstitium, which shows up here since it the blood serum was labeled by a red dye as this burst of of, dye that flows out into the interstitium.

Finally, through, more intravital imaging of these, Tmem structures, we’re able to go and see that when that door opens, other tumor cells are able to flow into the vasculature and then spread hematogenously.

Here you can see, all of those tumor cells which wind up flowing in. And so this is the moment when the tumor cells actually entravitate and spread to secondary sites. And so all of this research has really shown us that, TMEM are in function as doorways for other tumor cells to intravasate and spread to secondary sites like the bone or the brain or the lung.

Now we ask the question next, what’s so special about these motile cells? And in order to answer that, we developed another assay that we call the in vivo invasion assay, which is the only assay that can separate those motile cells from the rest of the bulk tumor.

What the assay consists of is a a a set of artificial blood vessels, what we call artificial blood vessels, and these are just microneedles that are filled with a growth factor that can attract those motile cells, which approach the needle and actually crawl up inside of the needle. After that, you can take the needles out, expel the contents onto a petri dish, and then do, analyze them with other other methods. So when we did that, we did expression profiling of those motile cells, and what we identified through that was the protein called MENA, which is an actin regulatory protein, and we found that MENA is very highly upregulated in those motile cells compared to the rest of the bulk tumor.

Now MENA is a protein that has a number of different splice isoforms, alternatively spliced isoforms. And some of those isoforms, one in particular that’s called MENA eleven a, is highly associated with epithelial phenotype, kind of like a normal, breast architecture. Whereas the others are associated more with mesenchymal phenotypes that when they are overexpressed in the cell that, the cell becomes highly motile and migratory.

And so we’ve developed this metric that we call MENA Calc, which measures the fraction of those MENA isoforms that are specifically associated with metastasis, and therefore, we can put a number onto that tumor that describes how much those tumor cells inside are able to migrate.

And so taken together, these new findings really presented us with a new model for metastasis.

One wherein some of the tumor cells inside of the primary tumor become MENA Calc high, and they pair up with macrophages, and they start migrating together toward the blood vessels. Once they reach the blood vessels, they either assemble new Tmem doorways or they interact with Tmem doorways that are preexisting.

Those Tmem doorways then open up the door to the vasculature and allow these migrating tumor cells to entravitate, spread to secondary sites where they can grow into metastases.

And so we asked the question, can we use TMEM and MENA Calc, these two new insights, as biomarkers specifically for metastasis?

And in order to do this, we created two immunostaining protocols, both of them working on formalin fixed paraffin embedded patient samples. And this is really important because we wanted to be able to validate and and, show that these, biomarkers work well. And the only way to really do that is to, do this in a number of retrospective, samples retrospective studies where you take samples that had been excised from patients twenty years ago, and then you ask the question, how did they fare whether with when your biomarker is high or low.

And so the first of these, protocols is called we call MENA Calc, and and it’s a quantitative immunofluorescence assay that consists of three different antibodies staining for the pan MENA, which is an antibody that stains all of the isoforms of MENA, and MENA eleven a, which is again that epithelial associated, isoform of MENA. And MENA calcus then calculated as the difference between all of the isoforms or or you just take the all of the isoforms and you subtract away that epithelial isoform. And this is then a metric of how invasive those tumors are.

We additionally use cytokeratin in order to limit where we consider, where we take signal from the the MENA proteins to just the membrane and exclude the nucleus or anything outside of the tumor cells just because, the MENA protein is actually a membrane bound protein, as you can see in the images on the bottom.

The other staining protocol that we developed is a triple immunohistochemistry stain, again, on fixed tissues that just marks the, component cells of of tmem structures. So we use Pan MENA to identify all of the tumor cells and then CD31 for endothelial cells and CD68 for macrophages.

Once we have those stains, our pathologist can come in and see where all of these three cells lie in contact with each other, in direct contact with each other. And then the pathologist just put a circle down on top of all of the TMN structures and, the TMN doorways, and then they count up the number of circles in a a certain area of of tissue.

Okay. So now that we have these two stains, we needed to clinically validate these biomarkers. And so we did this in a number of of different studies. We have two for Immunocalic and several others for Tmamm, Then all of these, span, several thousand patient samples. And what we found is indeed the, TMEM and MENA CALC markers are prognostic for specifically for metastasis.

And this really makes them unique because most of the, tests that are clinically available, such as IHC four or Oncotype DX, these are are, measures of the expression of genes that are specific for tumor growth and not for dissemination potential.

And, indeed, when we compared, in a large cohort, the TNEM score of individual patients to their Oncotype DX score, what we found is that there was absolutely no correlation, of outcomes that, or or scores between the two, markers. And that really indicates to us that both TNEM score and Oncotype DX score are giving us distinct biology, and so there’s new information that TNEM score is is providing.

So with this, with this confidence now that these biomarkers actually do provide information and can prognosticate outcome for for patients, we asked how can we prepare this test for the clinic? Well, if you are going to, run this test as part of, you know, a by a small biological study, having everything done manually is just fine. But once you want to start to use this as an actual test for, for patients, you need a higher throughput method for, for doing this. So we we decided we’re gonna develop a high throughput analysis technique. And so we turned then to digital pathology for this high throughput out, automation.

And so the story that I’ll tell you today is focusing on the our efforts to automate TEM, which we just published last year in the journal cancers.

There’s a similar story, but that I can tell for MINA Calc, but it’s, due to the time restrictions, I’ll only be able to go into the one.

So for TMEM, when we wanted to automate the the process, what we wanted to do is we wanted to really mimic the pathologist workflow as closely as possible. We didn’t wanna completely throw it out and use some other techniques like artificial intelligence. We really wanted to stick to what the pathologist had shown, works well. And so for the the procedure that the pathologist had worked out was to cut and manually stain the slides, then to view those slides at high magnification, on their microscope. And then they would identify the regions of high vascularity in the tumor tissue, acquire and save ten high power fields of view from their microscope, and then score the images in Photoshop. And altogether, each case took the pathologist about an hour of time.

And we wanted to dramatically cut this down. So the automated procedure that we came up with, replaced the manual staining with, automated staining on an automate on a high throughput, auto stainer, and then we replaced the viewing and acquiring of images on a microscope with a digital whole slide scan. We then used software in order to identify and segment the invasive tumor from the stroma and to identify the macrophages and blood vessels within the tumor. And then finally, we identified the TEMM structures based on their morphology and their juxtapositioning.

And then finally, we would cut that tissue up into the equivalent of these high high power fields of view and rank and score the team in the top ten fields top ten ranking fields. All in all, we were able to reduce the amount of pathologist time from about an hour per case to down to about five minutes, mostly just for quality control.

And so in order to understand this whole process, I’ll walk you through it next, but in order to understand the process, let me just mention briefly what digital pathology is and how we are using it. So digital pathology uses digital whole slide scanning. It’s basically a machine that can go and scan the entire tissue by acquiring thousands of, high resolution images that span the entire tissue and then stitch them together to create one giant super image. It then creates downsampled versions of those images so that you can seamlessly switch between magnifications very much like you’re using Google Earth where you can view the world and then a country and then, somebody an individual house.

And so the real power of digital pathology is not just though in viewing the tissue, it’s that we can actually perform single cell quantitative analysis on any one of these individual tiles and then apply that analysis tissue wide. And one of the, software packages that we use to to do this is VIS by by Visiopharm. So what does the process actually look like?

Well, first, we start with the digital whole slide scan. So here’s the entire, microscope slide. We can then zoom in and zoom in and zoom in again until you get down to the single cell level. We did this on our, three d HisTech Panoramic p two fifty, and we scanned at, forty x, taking six thousand five hundred fields in really under three minutes.

Then we developed a a series of custom developed apps within, VIS from Visiopharm.

And, the first one was just to perform the invasive tumor identification and segmentation so we know where we want to apply our our future analyses.

And then we, analyze that tissue at very high magnification using color deconvolution to identify the stain, of all of the the tissue.

Using k means clustering, we could classify each in the the color in each and every pixel into one of the categories of blood vessels, macrophages, tumor, or stroma.

And at that point, we could use the morphometric analysis capabilities of Visiopharm in order to identify the macrophages and blood vessels that were in direct contact with each other and in direct contact with tumor cells around. And here is just a zoom in of one of those where you can see on the left, the identification of the tumor cells, the the macrophages and the endothelial cells, and on the right, what Visiopharm was able to identify in an automated fashion.

And then we apply that single analysis across the entire tissue. At that point, we can cut up the tissue and rank score and rank all of the fields of view, and then, the algorithm is able to present us with the ten top, ranked fields of view from which we can generate an overall score for the whole tissue.

So now that we have an algorithm and we have a process, a method for evaluating TM, we wanted to validate that that process. How well does it perform? How does it compare to what the pathologists were doing manually?

So we, again, turn, we’re now using this digital pathology for analytical validation.

And so in order to do this, what we did was we picked out several measures of reproducibility and accuracy and evaluated them in order to see how each step of that process was being performed. And of course, you know, the first thing that we wanted to do was to evaluate how well do pathologists, perform the TMEM identification.

And so what we did was we took a hundred test fields of view that contained a variety of of numbers of of TMEM doorways in them, and we gave them to two pathologists on two different days, and we asked them to to score them. And what we found is that within observer, right, so one pathologist on day one and day two, they agree with the themselves, to a high degree with ninety six percent of the time, on for both pathologists, they agree with themselves on the next day. But they also agree between themselves. So there’s a high level of of agreement ninety two percent of the time between pathologist one and pathologist two.

Next, we wanted to ask how well can the algorithm do a similar kind of thing. So we gave the all of the slides over to the algorithm, scanned them twice on two separate days, and we asked how reproducibly could the algorithm, identify those TMEM doorways. And what we find is, unsurprisingly, because it’s a machine, the day to day reproducibility was perfect. We’re agreeing with itself a hundred percent of the time. The machine does what the machine does.

Right? So the next step was to ask, how well does the algorithm perform in identifying a TMEM structures within a field of view that’s given to it? So you give the algorithm one particular image and you say, count up the number of of t men doorways in this, how well does it prepare how well does it perform doing that? So we gave we picked out ten field fields of view per case, that were selected by one pathologist, and we gave them to the machine and to an a pathologist another pathologist for team and scoring.

And what we found is that we also get a high level of, of, correlation between the scores that are generated. About eighty six percent of the time, they agree.

Finally, we wanted to test the whole procedure, and the whole procedure involves giving a pathologist the entire microscope slide and saying, there’s no restrictions. Go look anywhere you want on the slide and go pick out a field of view, score it for the t m score, and return me a result for the entire slide. And then we do the same thing for for the machine algorithm. And so we did this we actually did this with five different pathologists scoring, and then we compare the average of those five scores to the the machine generated algorithm. And what we found is actually an even higher level of concord of, correlation between the pathologists and the algorithm where they agree about ninety four percent of the time. And we believe that this is higher than the eighty six because we use the average of five different pathologist scores, and it seems like the the pathologist to pathologist variation seem to be driving down the score when we we just compare the individual fields of view.

Okay. So now that we have this validated test, it’s ready to go into the clinic. We decided we’re gonna go and see what what would be the next thing that we could do to move this, to improve upon the performance of these biomarkers.

And so this is our current work that we’re probably in the next week or so going to submit for publication.

It’s been the work of doctor Shunjun Yeh, who’s a postdoc a very talented postdoc in the lab. And, he was asking the question, can we actually improve that prognostication?

How would we go about doing that? And so the way to do it, the the rationale is you can get to the rationale by thinking about what these biomarkers are actually telling us biologically.

So the Tmem biomarker is a quantification of the number of doorways into the vasculature that there are inside of a primary tumor. And the MENA Calc is a quantification or measure of the level of cell motility for the tumor cells that are in there. But what you realize through this then is that if, if you have no motile cells, but you have lots of t m doorways, you’re not going to get a lot of metastasis because none there are no moving cells to pass through those doorways and and go to secondary sites. Similarly, if you have, no TMEM doorways but lots of moving cells, those cells can’t go anywhere. They can’t get access to the the vasculature.

And so in this case, no TMEM doorways means that you’re not going to have an appreciable level of metastasis.

Only when you have, both, motile cells and an appreciable number of Tmem doorways so that those cells can get into the vasculature will you get hematogenous spread and metastasis?

And so, we asked the question then, how can we effectively take these two mechanistically linked biomarkers and combine them in order to, to take advantage of this insight.

So a simple way, of course, to do this is to say we’re gonna combine, or call this combined marker high when both of the individual tests are high. And high is really determined by a cutoff threshold that effectively separates a a patient population, between a group that where they’re not going to have metastatic recurrence and a group where they are going to. And so for each individual score, for each individual biomarker, we can say that, Tmam score for a particular case is considered high if it’s above that cutoff and low if it’s below that cutoff and similar for, MENA Calc.

And the combined marker then would be high if both t men and mina cal for that particular patient is high and low in all other cases.

But before we can just go ahead with that simple, method, there’s another complication, which is that TMEM and MENA Calc, are measured, differently.

TMEM is actually, measured only in the top ten fields, whereas MENA Calc is measured across the entire slide.

And so there could be several different ways of combining TMEM and MENA Calc. They may there may be spatial associations. You may need to have, t men and MENA Calc measured in the exact same region or there may be no spatial connection is at all. In addition, we don’t know what the impact of tissue heterogeneity is. How much of that tissue do you need to acquire? Is ten fields of you enough, or do you need to go larger?

And finally, we want so so given all of this, we wanted to test a number of different ways of combining all of these markers.

And so what we did was we started out by defining what are the different tests that we’re actually going to to try out and evaluate. So we started by defining the extent of the tissue, which is really a way of accounting for tissue header heterogeneity.

We can either measure the entire tissue or we could have our pathologist limit that area to where they felt that would be, most representative of that tissue.

And then within each of those regions of interest, we could either analyze the number of team m doorways across the entire region of interest, or we could just pick out the top ten fields of view in each one.

So all in all, if you number these one, two, three, and four, it gave us four different tests that we could evaluate and compare.

And once we evaluated those, of course, now we have to mix in the mean account in those same areas.

And but there’s an additional consideration, which is that, prior studies using MENA CALC limited the quantification of the signals to just the cytoplasm by a cytokeratin mask. And it’s really never been tested whether this was necessary. So we additionally wanted to to test that out. And so given those four different areas and the two different possibilities for the cytokeratin mask, that leaves eight different combinations for MENA Calc. So we have four TMM analyses and eight MENA Calc analyses that results in thirty two different combinations, and that’s a lot of combinations.

So what we said though is that not all thirty two combinations are really consistent in ROI type and tissue coverage. So we wanted to only test those things that use the same, ROI type and tissue coverage for both, both individual biomarkers.

And so that paired the thirty two choices down to only eight, which use the same ROI type and tissue coverage. So how do we evaluate which one is best? Well, the the first step is to evaluate how each, individual biomarker, TMEM and MENA Calc, separate the patient population based upon time to distant recurrence.

And so how do we do this? Well, what we do is we adjust that cut point, which defines, whether a patient is in the high group or in the low group, and then we look at how they fared over time, using a Kaplan Meier analysis.

And what we see, when we do this is that the, Tmem and MINA Calc, tests alone, they work fairly well. They create, we can select cut points for all of them and get the p values, and then we can combine them using that, tech of the logical end operation that I I mentioned before to get the p values for the combined markers.

But it’s a little bit difficult to just tell whether there’s any improvement in, in, by by just looking at the p values. And so what we did was we developed a better metric, which is to convert those p values to fold changes.

And the way we do that is we take the ratio of TMEM to the combined marker or MINI CALC to the combined marker analysis to generate fold changes in performance gain. Right? And what you can see from this is that the combined marker almost always improves prognostication versus TMEM or the MENA Calc analysis alone.

And you can see that because all of the numbers are larger than one except for the TMEM four MC eight, which is a zero point five. So only in that one case does it not outperform the individual, biomarker.

And what you can also see is that you actually get a better improvement in general for MENA Calc than you do for tmem. The numbers are are typically higher for, fold improvements for the performance gains in, MINA Calc than for TNEM.

So which one is actually best? Well, in order to narrow it down, we decided to highlight, just those scores where t m the t m gain is greater than five and the mean account rate gain is greater than ten, and this yields two winners. We can see the t m one, m c two is a winner and t m four, m seven is also a winner. But if you look at the values, TMEM four MC seven has a full gain of eleven point three and six point nine for the two over the two individual biomarkers, whereas the other test, the TMEM one MC two, has a performance gain of of a hundred and seventy five and a hundred and eighty seven. So the best clear the clear best performer gives us a greater than hundred and seventy fold improvement over both TMEM and MENA Calc alone.

So how well does that best performer separate the population?

Well, you can see here that the TMEM and MENA Calc analysis are able to separate the populations, somewhat well. But when we evaluate the same cohort with the combined marker analysis, we can see a dramatic improvement in in segregation between those two separation between those two groups. So the combined marker analysis dramatically improves the prognosis of distant metastasis.

And what does that winner look like?

Well, the best performer analyzed TMEM over the entire tissue and excluded the nuclear MENA Calc signals by using the cytokeratin mass. And this really indicates to us that the tissue heterogeneity and nonspecific MENA staining are crucial factors.

So in summary, I told you about how we discovered these new biomarkers from metastasis and how, we did clinical validation of these biomarkers using digital pathology to show reproducibility and to analytically validate them, and then how now we’re trying to use digital pathology to improve prognostication.

And all of this work has really been made possible by the Grufflicker Biophotonic Center and its associated integrated imaging program at Einstein, where the mandates are to develop novel technologies and to use those novel technologies in order to bridge to the clinic. And what I’ve told you today is just one story that we have in collaboration with the department of pathology, but there are several others with department of surgery and our health care industry partners and collaborators both, across the US and abroad.

I’ll end end up with just showing you the clinical and basic science research team, which spans a number of different institutions that, work together under the auspices of the New York Pathology Oncology Group. And, I’ll finish up with just displaying my acknowledgment slide and take any questions that you guys have. Thank you.

And thank you, David, for that informative presentation.

We will now start the live q and a portion of the webinar. And joining David is Regan Baird, the regional manager for Visiopharm East. I wanna remind our audience members that if you have any questions you wanna ask, please do so now. Just click on that ask a question box located on the far left of your screen, and we’ll answer as many questions as we have time for.

So let’s take a look. Reagan and David, we have quite a few questions already coming in.

Our first question is the two tests that you’ve developed use different modalities.

TMEM is an immunohistochemical stain, and MINI CALC is an immunofluorescence stain. Why not make them IHC or Bose IF?

Great. Thanks. Thanks, for the question. So, yes. They’re both taken in different modalities, but they are fundamentally measuring different things about the tissue.

So, immunohistochemical stains are typically not very, quantitative in the stain. The the process of the development of the chromogen actually is a a pretty nonlinear process, whereas immunofluorescence, the amount of light signal that you get from a a from a fluorophore can be quantified and related directly to the protein expression level. And so, for me the calque where we’re really interested in the level of the isoform, expression, we will use the immunofluorescence.

However, for the tmem stain, what we’re really more interested in is the morphological, the morphology of the tissue, that is the juxtaposition of the three different cell types. So we’re really looking in that case more for a yes or a no answer to is this particular cell a a macrophage or is it a tumor cell, and what is it next to? And so that’s why, we really use these different, the two different modalities.

As part of our second study where we compared, the two biomarkers we use and tried to create the combined marker, it was really important since we were using those two different modalities to be able to, use some of the to align features within Visiopharm.

That, really could align those two different images, the the two slides down to the single cell level, and that gave us the ability to have confidence that we were looking at the same region, sometimes the same cell within the tissue, in order to be able to evaluate both the morphology of of the TMEM and the expression level of the isoform, protein isoforms.

Thank you so much. Now what is the hypothesis for the clinical value of the TMEM results for guiding the decisions of management of patient therapy?

Well, thanks. That’s actually another good question. Sometimes it’s not so clear, you know, why you would want to be able to specifically measure the level of dissemination. And, really, the answer is is because, metastasis and and, metastatic spread is really what is responsible for the overwhelming majority of mortality in, breast cancer.

The problem that clinicians face is that they don’t know who is going to have metastatic disease and who isn’t. And so pretty much everyone who, has tumors of a certain characteristics or over a certain size, or a certain grade winds up getting chemotherapy just because, we don’t know who who has metastatic disease and who doesn’t. And so that leads to, potentially to a lot of overtreatment of patients, and and these treatments can be very costly in terms of money. They’re expensive treatments, but also in terms of quality of life. They’re very harsh.

They are cytotoxic, therapies.

And so if we can reduce the amount of of, overtreatment, then that will have a a very significant impact on lives of of patients. In addition, there’s also a smaller subset of patients who potentially would have metastatic disease, but it’s not apparent from their primary tumor. And so the tests like MINI Calc and TMEM are really designed to to try to identify that particular population that’s going to have, disseminating disease and would benefit the most from, from cytotoxic chemotherapies and also, you know, to to identify those that would not benefit.

Thank you very much. And I wanna also thank our audience members for these great questions coming in. Our next question from an audience member, is there a difference in the Menocalc value obtained from different parts of tumors or different areas of the tumor that may be hypoxic?

So is the I guess the question is really about the heterogeneity of of the Immunocalc.

And part of what we tried to address in this, second study with the combined marker is what really what is the impact of tissue heterogeneity on the the the prognostic ability of this biomarker? And and what we’re finding is that, tissue heterogeneity exists, and, it is actually a bit detrimental to our our ability to prognosticate.

And so the solution that we’ve, believe we’ve struck upon with the results of our combined marker analysis is that the, is that, taking the meanicalc or averaging the meanicalc over the entire tissue slide is probably the best way to go. And that way, you you tend to, average out the areas that are heterogeneous.

And, areas like hypoxic areas, we’re finding in other studies actually, do have an impact upon both TMEM formation and MENA Calc expression, and so that these are these conditions of hypoxia wind up leading to increased levels of of dissemination.

Thank you very much. Now how does your combined biomarker analysis relate to local relapse and lymph node metastasis compared to distant metastasis?

So what we have been studying is, really geared toward, an analysis of distant metastasis.

Local relapse is best measured by, other metrics of that are more closely related to growth.

So for example, I showed the, comparison to Oncotype DX. This is a a a a gene analysis which, really tells about the growth potential. What are the cell proliferation genes that are regulated inside of the tumor? And that would wind up leading to local relapse. The difference between lymph node metastases and distant metastases, it it’s not clear.

But an additional complication to that is that, it’s been recently published by our group and by other groups that dissemination hematogenous dissemination can also occur from secondary sites like lymph node metastases.

And, in particular, in a study that we published just last year, what we discovered was that, TMN doorways actually exist, can be visualized inside of of lymph node metastases.

And so, the presence of tumor metastases inside of lymph nodes could, be an indicator of of further spread, and, that’s due to to TMEM. And so by being able to, assess the level of of TMEM doorway and Immunocalc involvement in the primary tumor, it could potentially inform you about, the dissemination, confidence of of of, lymph node MET as well.

Thank you so much. And we have time for one more question. Are these biomarkers specific only to breast tissue?

So no. They’re not specific, only to breast tissue.

I believe that, right, so we’ve done most of our studies inside of breast tissue, but we believe that the, TMEM and MENA CALC biology is actually common to all solid adenocarcinomas.

And we have been able in our studies to see this in lung adenocarcinoma, pancreatic, succuladenocarcinoma, breast cancer, and, there’s one more that’s, Ewing’s sarcoma. I’ve even, we’ve even observed these TEMM doorways. And and so, the studies that we have going on, in fact, there are even a couple of clinical trials, looking at inhibitors of of, TBM dissemination in a variety of different cancers. And so we we feel that this prognostic that we’ve developed should have utility for a wide variety of cancers.

David, I wanna thank you again for your time today. Would you like to provide some final comments for our audience members before we go?

Sure. Yeah. So I I just think, that, the type of work that we’re doing in addition to being, you know, fascinating and and interesting, has the potential really to to, make an impact upon patients’ lives. And so, we’re really excited to be able to, bring these to fruition and, to partner up with companies like Visiopharm where, a lot of times, our initial idea sounds crazy and it’s somewhat impossible to do, but if you go and find the right partners, then it it actually makes it possible.

Thank you so much, David. And thank you for your time and for your important research, and thank you for joining us today, Rian, for the q and a. I’d also like to thank LabRoots and our sponsor, Visiopharm, for underwriting today’s educational webcast. Before we go, I wanna thank our audience for joining us today and for their interesting questions And questions we did not have time for today and those submitted during the on demand period will be addressed by our speaker via the contact information you provided at the time of registration.

And this webcast can be viewed on demand. LabRoots will alert you via email when it’s available for replay. We encourage you to share that email with your colleagues who may have missed today’s live event. Until next time.

Take care. Stay healthy. Stay safe. Bye bye, everyone.

K. Thanks so much.