Well, good day to everyone joining us, and welcome to today’s xTalks webinar. Today’s talk is entitled Strategy for Visualizing, Quantifying, and Mapping Immune Cells in the Tumor Microenvironment.

My name is Edward, and I’ll be your Xtalks host for today.

Today’s webinar will run for approximately sixty minutes. This presentation includes a q and a session with our speaker. This webinar is designed to be interactive, and webinars work best when you’re involved. So please feel free to submit questions and comments for our speaker throughout the presentation using the questions chat box, and we’ll try to attend to your questions during the q and a session.

This chat box is located in the control panel on the right hand side of your screen. If you require any assistance, please contact me at any time by sending a message using this chat panel.

At this time, all participants are in listen only mode. Please note this event will be recorded and made available for streaming on xdocs dot com.

At this point, it’s my pleasure to thank Visiopharm who developed the content for today’s webinar.

Visiopharm is a world leader in AI driven digital precision pathology software.

Leading biopharmaceutical companies, CROs, academic medical centers, and diagnostic pathology labs all over the world use Visiopharm’s technology for tissue based research and diagnostics.

Its solution use the latest advancements in artificial intelligence and deep learning to make the most comprehensive, highly configurable, and accurate tissue mining tools available on the market today.

Visiopharm was founded in two thousand one and is privately owned. The company operates internationally with over nine hundred licenses in more than thirty eight countries.

Company headquarters are in Denmark’s Medicon Valley with further offices in London, England, Munich, Germany, and Westminster, Colorado.

And now without further ado, it’s my pleasure to hand the mic over to our Visiopharm host and our moderator today, Regan Baird.

Regan will be introducing our speaker. Regan, you may begin when you are ready.

Hello. Good morning and good afternoon.

Thank you for joining today’s xTalk webinar hosted by Visiopharm.

My name is Regan Baird, scientific consulting sales manager at Visiopharm.

Today’s speaker is Manuel Flores, PhD candidate at Tashoum University de Montreal in the laboratory of doctor Nagla Shokery.

Manuel Flores obtained his biochemistry degree from Havana University in Cuba. In two thousand fifteen, he enrolled in the immunology and virology master’s program at Universite de Montreal and fast tracked into the immunology and virology PhD program in two thousand sixteen.

He is focused on characterizing the liver resident and infiltrating immune cell populations and their role in pathogenesis of chronic liver diseases due to persistent viral and toxic injuries, including fibrosis and hepatocellular carcinoma.

His research interests center around the spatial organization of immune cells in the hepatic tissue microenvironment and the delineation of the multiple cell to cell interactions and their respective biological significance in health and disease.

Manuel is the recipient of doctoral scholarships from the University of Montreal and from Fonds des Research du Quebec, Santee.

The title of today’s presentation is a strategy for visualizing, quantifying, and mapping immune cells in the tumor microenvironment.

We look forward to this presentation, and we thank you for your attendance.

Please stick around for the question and answering session to follow.

Hello, everyone. Thank you for joining this webinar, and thanks a lot to Visiopharm for this opportunity to present our work to its audience and for all the conceptual and technical support to our projects over the years.

The title of this presentation is a strategy for visualizing, quantifying, and mapping immune cells in the tumor microenvironment, and it’s based on a technical article we published, last year.

So here is an overview of the presentation.

I will first provide a brief context, on the topic of the presentation, then I will introduce a strategy that we put together combining in a rational way several simple solutions, for multiplexing that people have been using during the last years with tools created by Visiopharm that maximize the number of markers that can be visualized and analyzed simultaneously. Finally, I will walk you through one example where we applied this strategy and generated a visual slide where we visualized ten immune markers plus DAPI, contrast stain, and two other frequent stains.

And we were able to quantify these biomarkers in different tissue compartments and establish how these parameters especially relate in the tissue.

Why do we need to study the immune response in the tissue context?

Immune responses are spatially and temporally regulated.

This is very clear when we look at the impressive compartmentalization of secondary lymphoid organs with specialized regions like T cell zones, B cell zones, germinal centers, etcetera.

But the same is true to some extent for any given non lymphoid tissue. The fact is that resident immune cell populations are strategically positioned, they occupy specialized niches, and they establish stable and dynamic interactions with neighboring tissue cells and the extracellular matrix.

And one hallmark of tissue pathology is precisely the disruption of this spatial organization and the composition of the immune compartment in the tissue.

Therefore, it is not surprising that the special organization of immune cells in the tumor microenvironment has prognostic and diagnostic values.

The most remarkable clinical translation of this idea is the immune score in colorectal cancer.

The immune score is based on the quantification of memory t cells and memory cytotoxic CD80 cells in the center of the tumor and in the tumor margin.

The immune score has been shown to outperform the gold standard TNM classification system in colorectal cancer, demonstrating that information on the composition, density, spatial organization, and functional orientation of immune cells in the tumor microenvironment can help stratify patients and predict response to therapy.

Finally, in the recent years, the emergence of technologies allowing single cell multiomics profiling has exponentially increased our understanding of the factors regulating carcinogenesis and tumor progression.

However, most single cell based technologies require tissue disruption and single cell isolation, resulting in loss of information about the spatial organization of cells and cell cell interactions, and potential misrepresentation of specific cell populations that are not efficiently isolated.

On the other hand, there are few technologies, very promising, that allow highly multiplex tissue imaging analysis, like histocytometry and imaging mass cytometry.

But they use reagents, equipment, and software that are not readily accessible to the majority of the researchers. So there is a need for the development of affordable and accessible imaging, imaging analysis techniques, and software tools that increase the multiplexing capability of imaging assays.

Some of the major obstacles in the way of developing, such assays are mentioned here. For instance, immune cells are typically defined by multiple markers, but the microscopes, most of them equipped with very limited number of placers and channels, and consequently, visualizing several immune populations simultaneously can be very challenging.

Fluorophores stable enough for imaging are also limited if you compare with flow cytometry, for instance. In addition, multiplexing with conventional technologies requires the primary antibodies to be raised in different species or being of different isotypes, and this is frequently impossible to achieve with antibodies that are commercially available.

However, researchers and technicians have found ways to circumvent these obstacles through serial imaging, sequential labeling, spectral imaging, tissue alignment, and virtual multiplexing, all of them solutions that are widely accessible.

In our lab, few years ago, we started imaging immune cells in liver biopsies and resected tumor sections. So we designed a methodology, summarized in this slide, that integrates easy to use and well known techniques in a workflow that maximizes the spatial information that can be obtained from precious clinical specimens.

The first step in this methodology is imaging the cell populations and tissue compartments of interest, meaning that you need to define what is the cell population you want to characterize and in which tissue compartment.

To do this, we use serial sections, and when possible, reuse the section through sequential labeling.

The images are acquired in a whole slide scanner, then aligned, and this is the second step in the strategy.

They are aligned using the tissue aligned module of ViS. This is going to generate a virtual slide containing all the information of the individual images.

Next, we detect the tissue associated pixels using one of the aligned images and create a ROI for the tissue.

Next, the tissue is segmented, in into tissue compartments, convenient tissue compartments, like a stroma, parenchyma, tumor, peritumor.

And next, the cell populations, of interest, are quantified in these tissue compartments.

The last step in the strategy is the generation of tissue heat maps of the cell population that have been characterized.

I have to say that, this methodology is implied in the educational videos of Visiopharm .

In fact, Visiopharm’s has devoted, several webinars to individual steps that are included in this strategy.

But we thought that putting all the pieces together in a single strategy or methodology would be very helpful as a guidance or researchers that are starting to plan imaging analysis. So people with no experience of imaging, they can have a glimpse at all that they can extract from the tissue by looking at this, strategy.

Now let’s see this strategy applied to a concrete example.

Three serial formally fixed paraffin embedded sections from resected hepatitis b virus associated human hepatocellular carcinoma were stained as follow.

Section one was exclusively used for h and e staining. As you know, this is a a stain that is routinely used to determine clinically relevant parameters. It was not possible to reuse this section.

In the consecutive section or section two, two rounds of multiplex immunofluorescence were used for labeling liver parenchymal and nonparenchymal cells. In the first round, tumor vessels were visualized using c thirty four.

Epithelial cells like hepatocytes and cholangiocytes were identified using cytokeratin eight eighteen, and fibrogenic activated hepatic stellate cells were identified as alpha smooth muscle acting positive cells, alpha SMA positive cells.

Following image acquisition, the antibodies in section two were a strip, and the section repro with antibodies against macrophages, the marker we used was CD sixty eight and myofibroblast, and the marker we used was desmin.

Finally, for section three, we did three staining.

First, we did two rounds of immunofluorescence for identifying different immune populations, like, regulatory FOX p three positive T cells, CD four T helper cells, MPO positive neutrophils, CA three total T cells, CDA positive cytotoxic T cells.

And then the first staining was the pico serious red staining. It’s a very common stain in hepatology because it allows you to visualize, the fibrillar collagens. In our case, we use this staining for tissue segmentation into a stroma and parenchyma.

In all cases where when we did immunofluorescence, DAPI was used as a nuclear contrast stain. In total, we generated six images out of three tissue sections.

The selection of markers was based on the performance of the antibodies, on the fact that we wanted to have both immune cells and tissue cells labeled, and also that we wanted some staining that would allow us to do tissue segmentation.

For sure, these markers can be easily replaced by any other phenotypic or functional marker of interest.

So the six images were imported to the VIS tissue alignment module and aligned, as seen here.

Upon the alignment, a composite image is generated that contains all the layers corresponding to the six individual images.

This is a sum of the, high magnification view of the different stanings, after alignment.

I wanted to show you the quality of the staining and also to show you the effect of the stripping on the quality of the staining. Here you have on the left the image before the stripping and then the image after the stripping. We found that the stripping procedure, makes the DAPI signal more diffuse. This is what you see here comparing these two images. Other than that, you can see that the stripping is very efficient at removing antibodies from the previous round of labeling and does not interfere with the quality of the second round of labeling as seen here and also as seen here. You see that the image on the right, has also high quality, and this is, an image after the stripping.

You can also, appreciate that for some stains, like the Pacrose Serious Red staining, they can be done with good quality on a slide that have been previously used in multiplex immunofluorescence.

And this is also true for immunohistochemistry.

We tested this protocol when we did two rounds of multiplex immunofluorescence, stripped antibodies, and then did immunohistochemistry.

And if the antibody is a good antibody, you will find that the staining is high quality. You can use it for quantification, for tissue segmentation, or for any other, imaging purpose.

We validated the quality of the alignment. Here, you can see on the on the left panel.

The alignment was accurate in the case of images originating from adjacent serial sections. In this case, the c thirty four and the H and E stain are two serial sections.

The green mask generated using the c thirty four staining almost perfectly match the vessels in the h and e slide. I’m aware that it’s very hard to see the vessels in the h and e slide, but if you make an effort, you will see them, and you will see a very great accordance with green c v thirty four, mask.

The alignment was also or even more precise for images originating from the same section. You can see a perfect match at individual cell level when the FOX p three mask in magenta is overlaid on the other image coming from the same section, the c u three, c d eight image. So this is demonstrating that really the tissue alignment tool of Visiopharm can reach single cell level accuracy. So it’s very powerful, this tool.

The next step in the strategy is identifying the tissue. So once the image were were aligned, we identified the tissue associated pixels. We use one of the immunofluorescent images, the one with the cytokeratin, but any of the other aligned images could be used instead of this one. We took advantage of two properties that differentiate the tissue associated pixels from pixels that do not belong to the tissue. First, the dAPI signal, blue band, is restricted to the nuclei, which are located exclusively in the tissue.

Secondly, the hepatic tissue has considerably high autofluorescence in the green and yellow bands compared to pixels not associated with the tissue. Consequently, we developed a simple app. Here it is.

And for this app, we use, as a classification method, thresholds. We set threshold for the blue, the green, and the yellow ones, and the app essentially selects the pixels with above threshold values as tissue and labels them in green. And then this label is subsequently converted into a region of interest during the post processing steps. So the resultant region of interest applies to all the aligned individual images. You can see it here.

So from now on, all the pixels that are outside this region of interest are excluded from ulterior analysis.

Next, we proceeded to define different compartments inside the region of interest tissue by cementing this region of interest into stroma versus parenchyma. We used the picroserios red stain image, where the stroma was defined as the red area associated with the fibrillar collagens and the parenchyma as the green area where the fibrillar collagens are absent.

This protocol required training of the classifier.

You can see here.

And we use the decision forest as the classification method.

There are also a number of post processing steps in this app that are meant to better delineate the stroma and the parenchyma.

And as output variables, we get the areas of the region of interest for the stroma and the parenchyma. So we get the area of the stroma and the total area of the parenchyma. I want to mention here that for this app, we adapted from a an app that was developed by Michael Persch from Visiopharm, which we greatly appreciate.

So here you see the results of running this protocol on the predefined probe for the tissue. At this point, the stroma and the parenchyma are separated regions of interest. And even though the cementation is done using the PSR stained section, the aligned stroma and parenchyma regions can be transferred to any aligned image as seen here.

So next step in strategy is the quantification of the cell population of interest.

On the left, you have the actual images, and on the right, you have the respective process image and the quantifications.

I have to mention that the quality of our WAP staining was not good enough for doing nuclei segmentation.

So since we could not ensure that all individually labeled objects were individual cells, we expressed the density of cells in counts of labeled objects per square millimeter.

However, cell aggregates were successfully separated into individual cells in the processing steps in the different protocols. And by visual inspection, we verified that most labeled objects corresponded to single cells.

As seen here, we determined the density in the stroma and in the parenchyma of CD4 FOXP3 double positive cells identifying Tregs, CD8 T cells, CD68 positive macrophages, MPO positive neutrophils, and also the percentage of positive area in the case of c thirty four and alpha SMA.

Next and final step in the strategy is the generation of tissue heme map. This is a very useful tool of the Wyss software. I brought here a visual example for people not familiar with tissue heat maps. First, the cells of interest, for instance, these white cells on the left image are labeled. In this example, they are labeled in magenta. Now that you have an image with labeled objects, a heat map can be generated. The protocol for tissue heat mapping divides the whole image in area units of a predefined size and classifies these area units in red hot spots or blue cold spots according to the high or low density of labeled objects inside this predefined area, as illustrated in the image on the right.

So this is how the app for tissue heat mapping looks.

I I guess the math and the coding behind this app, like everything in this, is pretty sophisticated, but on the side of the users, it is very little what we have to do. This app has no classification, no post processing. The only thing we need to do is adding the processing step object heat map in the feature sections. As you can see, all the user needs to specify is the label to be used for the heat map, the object measure, which is either the count or the area, and the drawing radius that defines the area of the circle that is used at the unit of analysis.

Coming back to the strategy, here are the heat maps for FOXP three CD four double positive cells or Tregs, CD eight T cells, CD sixty eight positive macrophages, MPO positive neutrophils, CD thirty four positive endothelial cells, and alpha sema positive activated hepatic stilate cells. These heat maps were generated based on the labels for the whole tissue without segmentation into stroma parenchyma. This is important to mention. As you can see, tissue heat maps provide a panoramic view of the spatial distribution of a given cell population or biomarker, revealing hot spots or region where specific populations are depleted or absent. Tissue heaps can be very informative.

For instance, just by looking at these images, you can tell right away that these cell populations are not distributed homogeneously in the tumor microenvironment.

In fact, there is impressive heterogeneity for all the biomarkers visualized here.

In this slide, on the left, we have a line with black dotted lines. The nodules is identified by a pathologist, four nodules. On the right, you can see in every column one of these nodules and the associated heat maps for the different immune populations. As you can see, it is remarkable that even in the same section from the same patient, this nodule have a very different immune signature, and this can be revealed by the tissue heat maps in a very obvious way. Nodule one, for instance, has very few FOXP three CD four double positive Tregs compared to nodules two and three and four.

All the nodules share exclusion of CD80 cells from the tumor. Nodule three and four are heavily infiltrated by MPO positive neutrophils compared to nodules one and two. All this varied information can be obtained by only looking at the tissue hemlaps, So it’s it’s impressive. It’s very useful.

In fact, we quantify different immune cells in these nodules as seen here on the right, and the results reveal that every nodule has its unique immune signature. In some way, each nodule in the same patient from the same section has its unique tumor microenvironment, which is absolutely consistent with the visual information provided by hit by the heat maps.

So now I want to mention some advantages of the strategy. Some of them are that, it is applicable to clinical specimens and and maximizes the information that can be collected from limited samples.

Imaging techniques like serial and sequential labeling are accessible and easy to execute.

The use of the VISS software requires no coding or programming skills from users, incorporates multiple solution and tools like tissue alignment and tissue heat mapping.

In this, you have user friendly design of apps or protocols allowing user specific customization.

The use of software tools for the automatic quantification greatly simplifies and accelerates the processing of images and reduce the subjectivity associated with manual quantification by visual inspection.

So just for any of this population we analyze, for instance, the Tregs, in this section, you will have thousand and thousand of Tregs. It would be impossible to quantify that manually by visual inspection. So using software tools really accelerates and increase the extent of the analysis.

Finally, this analysis can be performed on whole tissue sections instead of selected fields of view, resulting in an unbiased representation of tumor microenvironment.

On the side of the limitations, I will say the most important one is the fact that we couldn’t do nuclear segmentation. This is very important because individual cell delineation facilitates downstream identification and quantification of biomarkers and allows more accurate comparisons and variation with our techniques like flow cytometry.

However, nuclear segmentation could be easily incorporated into this strategy if you have the right app for it.

For this strategy to work, multiple techniques need to be optimized, like the different staining and the sequence they are performed in, as well as the antibody elution protocols.

In addition, several antibody combinations may need to be tested before finding the ones that work.

Finally, since users can design their own apps, it is important to validate them properly, and this is time consuming and also difficult to do.

As a conclusion, we have provided a strategy that maximizes the quantitative and spatial information that can be obtained from valuable clinical tissue samples.

The resources, equipment, and knowledge required to implement this methodology are widely accessible.

And that’s all. I would like to, thank first all the co authors of this study, Thomas Fabra, Aurelie, Genevieve Susie, Lilian Menier, Mohammed Adenavi, Nicolas Belforte, Simon Tucotte, and my supervisor, doctor Nagla Shukri. I also would like to thanks Visiopharm, especially Michael Persch, Dan Hori, and Rigambar. They gave me a lot of technical help and guidance.

And also the organizers of this webinar, Rainer Ganyos, Bettina Winkler, and Dana Murphy.

I also would like to acknowledge the study participant, the Biobank, and Luis Rousseau, who is in charge of the Biobank, the molecular pathology platform at the CR soon. The person in charge of the visa station that helped me a lot, Isabel Clement, and the funding agencies, Canadian Liver Foundation, their FRQS AIDS, and the Gang Hep C. And thank you to all of you for your attention.

Thank you as well, Manuel, for that very insightful presentation.

At this point, I’d like to pass the mic over to our Visiopharm host, Regan, who will be conducting our live q and a. For our live q and a today, I’d ask that our audience members please send in their questions via the chat panel on the right hand side of their screen. Thanks so much. Regan, take it away.

Thank you, Edward.

So we’ve got a few great questions that have come in already. Please keep them coming in, and we’ll try to answer as many as possible.

If there are questions that we can’t get to in today’s session, we certainly will respond back via email after the session.

Manuel, the first question that came in is a good one. It says, the tissue alignment of individual images generates virtual multiplex slides.

Other than for visualization purposes, can you do cellular colocalization between the different slides in the multiplex?

Thank you, Reagan, for this question.

As I show, when the images, align are derived from the same section, the alignment can achieve single cell level precision. In this case, you can analyze the colocalization biomarkers.

This is very useful when you have two antibodies raised in the same species and it is not possible to use them in the same cocktail.

In this case, you can generate two images from the same section using sequential labeling, generate a visual slide, and analyze the colocalizations.

In fact, we did this for a Desmond and Alpha SMA in our, article, but I didn’t introduce I didn’t present this in, now in the presentation, but it is possible to do.

The the condition would be, high precision alignment, and this can be easily achieved, with, the this alignment tool.

Yeah. I don’t know if that answers the question.

I think it’ll suffice.

Another question.

Yep. You know, we know that there are really impressive cell and nuclear segmentation routines available on the Visiopharm platform, but the quality of the DAPI staining is critical for the nuclear segmentation.

Is there any way that you could prevent the DAPI diffusion that you’ve observed during the antibody stripping protocol to get better pictures of the nuclei?

Yeah. There is. Unfortunately, we realized that, too late for this article.

The problem in our case was that we were using a DAPI that is, included in the mounting media, and this DAPI, tends to diffuse more easily than the regular DAPI that you do ten minutes, five minutes, fifteen minutes, staining, and then you use a separate immune mount, media. So I think, I would recommend for people that is going to somehow, follow this strategy, to do, their, their DAPI staining using, not using the DAPI rays including the mounting media. This was our mistake, but, for this paper, it’s it was already too late. Yeah.

So we look forward to another, round of this type of experiment from you in the near future.

Yes. We are, this was just defining the strategy, and now we are applying this strategy to several projects. So, and we have no more this program with DAPI. Also, we know that, there is this, new app available in in in this, which is, an app that is going to, identify the nuclei by, artificial intelligence and deep learning. So, this is something I think is available for everybody and is very helpful. So we are using this for our current projects.

So now that you’ve learned how to do this dating method, how many rounds of labeling are you limited to?

There are a lot of stripping buffers available, commercially available, but we use a stripping buffer that is the one that is normally used for stripping antibodies from western blot membranes. This we tested three stripping buffers, and this was the best one that used SDS and beramarkaptotanova.

Everybody can, prepare this buffer. And using this buffer, we found that, three rounds was the top because after every round, you get an increase, in the autofluorescence in all the channels, and you need to make sure that your specific signal is several fold, the signal of the background, so you can easily identify your, your signal. So we found that three was a reasonable limit for us, but maybe using another stripping buffers and another stripping methods, you can, expand extend this to more rounds of of, labeling and stripping cycles.

Do you see a deterioration of the tissue as you do more rounds, and how does that deterioration affect the coregistration?

Well, there there is a because these are whole tissue sections, there is always some, part of the tissue that, suffer because of the manipulation.

So what, you can do is simply, check that your tissue, the integrity is preserved. If it is not, you have to exclude these areas from your analysis. This you can do manually.

And this is very easy to do, and then, you can proceed. But it’s very important checking that the integrity of the tissue is preserved and if it is not, corrected accordingly.

And then the follow-up question to that is, how do you validate and optimize the algorithms that you’ve developed?

Well, we have no formal training, but, as I mentioned in one of the advantages of this strategy is that, the design of apps in in in this requires no, programming skills. So you can, design your app yourself, and then, we do that, and we always check by how the app performs compared to a manual, manual manual accounting by visual inspection.

We analyze in in multiple field of view just to get an idea how the app performs in different part of the tissue because these whole tissue sections can be very different from one area to another area. And the way that we essentially, optimize the app is, by playing with three parameters. One is the magnification. So, we try to work at the minimum magnification possible that is gonna give us the data that we want to extract because increasing the magnification, reduce the speed of the app.

We also play with the filters.

Some filters are going to speed up your, your the the processing. Others are going to make the processing slower, and we also try to reduce to the minimum the number of post processing steps. These are the three principles we follow, and it has worked well for us. Our sections are very big. For instance, this one that I presented to you has three point two centimeters, so it’s a huge section. In this case, it’s very important that your app is, efficiently working because if not, you are gonna spend a lot of time doing simple fix like a tissue, detection or tissue segmentation.

Is your method currently available publicly?

Well, this is a published article. I I mean, I don’t I don’t know if I understand properly, the the question.

So the method’s already published so people can Yeah.

It is it is already published. And, with this article, what we wanted was to, provide a guidance to people that is, starting to do this type of assays. So, I I it’s not for advanced users. It’s for people that wants to, do some imaging analysis, and they don’t know all these different steps, and they may start doing their experiments without realizing that they need to plan in advance for, getting a benefit. And so it’s just to guidance for people that are starting to do, imaging and imaging analysis.

Well, that was fantastic. Thank you, Manuel. I think for the rest of the questions that are lingering, we’ll have to get back to our listeners via email.

Thank you very much, Manuel and Regan, for that lively q and a session.

We have reached the end of the question and answer portion of this webinar. If we couldn’t tend to we if we couldn’t attend to your questions today, the team at Visiopharm may follow-up with you after the webinar. If you do have further questions, please direct them to the email address on your screen. That’s marketing at Visiopharm.com.

Thank you everyone for participating in today’s webinar. You will be receiving a follow-up email from Xtalks with access to the recorded archive for this event. A survey window will be popping up on your screen. Your participation is appreciated as it will help us to improve our webinars, And I’ve just sent everyone a link in their chat box.

You’ll be able to view the recording of this event on this link, and you can also share this link with your colleagues. We encourage you to do that, and they will receive the recording as well when they register for the webinar.

Now please join us once more in thanking our Visiopharm host and our speaker today, Regan Baird and Manuel Flores. Thank you both again for such a wonderful presentation.

We hope you all found this webinar informative. Have a great day, everyone.