Hello, everyone, and welcome to the second of a three-part webinar series on biomarker testing in non-small cell lung cancer. Today's webinar will focus on tissue collection for biomarker testing. Next slide. I'm on the left-hand side of the screen. My name is Adam Fox. I'm a pulmonologist and clinician scientist at the Medical University of South Carolina in Charleston, South Carolina. And it's my honor to introduce our speaker for today's webinar, Dr. Nicholas Pastis. He's a professor of medicine at The Ohio State University, where he's the director of interventional pulmonology. He's also an indispensable clinical and research mentor for me personally, as well as many others, and served as my program director during fellowship training. As a final note, please submit questions in the chat and in the Q&A section. We'll be monitoring those, and we'll address those at the end of the presentation. Without further delay, I'd like to turn this over to Dr. Pastis. Thanks a lot, Adam. I really appreciate it. I want to thank CHESS for having me here. This is an honor to give this discussion part of the webinar. For disclosures, I've served as a consultant for Olympus, Cook, and received some research support from Philips. So why are we here today? We're going to talk about a few major aspects of EBIS, in particular as it relates to biomarker testing. We're going to go over some of the best practices, a little bit of a work in progress, a moving target, and those practices have to balance diagnosis, staging, and getting enough tissue at the same time. What's the role of EBIS in biomarker testing? Success rate for different modalities as best we can with available data. What's the role of rapid on-site evaluation? And then I want to address, or at least make the audience recognize some of the challenges of EBIS tBNA for biomarker testing. So really, this is just a nice picture to start with. This is the EBIS bronchoscope in the lung. On the left side of the screen, you can see the different lymph node stations, and really the point here is that the EBIS bronchoscope is probably one of the most important inventions in the pulmonary medicine in the last 20 years. You have almost complete access to the mediastinal and hilar structures for diagnosing and staging cancers and diagnosing other disorders. You can see a couple areas that the EBIS bronchoscope doesn't get in blue, perisophageal nodes, inferior pulmonary ligament, left adrenal gland that you can get with an endoscopic ultrasound, which can be complementary. But really, the workhorse is primarily the EBIS bronchoscope, particularly in the U.S. There are these couple stations you can't get, but oftentimes it may not make clinical difference if you can get the majority of those non-small cells that have 8 and 9 will have a disease elsewhere in the mediastinum that EBIS can get. So that's one way around it, although it can be complementary. You see a little diagram of the needle and the relationship of the bronchial wall and the ultrasound image. In general, what are our goals? We want to make a diagnosis. You can see on the top, you see a picture of what a histology smear is on the left. On the right, you can see a cytology smear, cells smeared on a slide. So first, is it cancer? And if so, is it non-small cell or small cell? Small cell has a different staging system, as you know, but in the realm of non-small cell, what type? And that's not enough now. That used to be enough 20 years ago. Now we want to know, is there a mutation in non-small cell? We want to know PD-L1 score, as those are potentially patients that could get targeted therapy or immune checkpoint inhibitor therapy. So just stepping back for a more global view, Adam's seen this slide. He was at a recent biomarker summit. This is a cartoon that was shared by Farhoud Farja, who is a thoracic surgeon at the University of Washington and a health sciences researcher like Adam Fox. This is a simple graph, but says a lot. So when patients come to your clinic, oftentimes they're referred for a biopsy of a mass. It's a little more complicated than that. You have to balance not just the diagnosis, which is what the needle biopsy, the knee jerk will get you of the mass, but what's the stage? And then not just the stage now, but are there any targeted therapies? What is the PD-L1 score? So your decision to biopsy has to factor all those three things. And on the same scale weighing, you have to weigh risks. Many of these folks are frail, and you don't want to put them through a needle biopsy of the mass, then to go back and do another bronch and biopsy the lymph nodes or liver biopsy when one test can accomplish everything. So you're trying to find that sweet spot, balancing all of these things into one decision. So I tried to think of one slide to kind of put it all together, and sometimes a picture is worth a thousand words. So this is a patient, and Adam sees this a lot, they're referred for an E-bis. When you have a hammer, everything's a nail, but maybe you don't have to do an E-bis on this patient. So let's kind of go through these CT fused images. So you can see a big hyler mass here. You can see a lot of adenopathy. You can see some FDG uptake along the pleura. So there's a high likelihood of a malignant effusion there. You can see the liver. There's several masses in the liver. And this one, I have two slices through the liver showing, again, part of your decision making. So sometimes you get the sense something's necrotic. See, there's no uptake in the middle here. So sometimes you're tempted to biopsy something bigger when maybe something smaller would give you more material. So these are the kind of decisions you might have to make. And then you see in the pelvis, you see in this vertebral body, you see sclerotic lesions with high FDG uptake that look like osseous metastasis. So it's not just, if you had to pick one test, E-bis may not be the best. That would be an N2 node right there next to that mass. So that wouldn't give you the highest stage. Remember, you're trying to complete staging if you can, at the same time balancing risk. So really a very low risk procedure here that would give you stage four disease, M1A, would be a thoracentesis. So that would be the procedure of choice here, not the E-bis. So I just want everybody to keep that in mind as we progress through this talk on E-bis. And then you may ask, well, why don't you just do these sites of distant disease, these bone lesions? Those can be biopsied if there's no alternative lesions. However, the decalcification process often makes them void of good DNA for molecular testing. So you have to keep all these things in mind when you're a person in a clinic seeing patients for staging. So back to genomic profiling, molecular analysis, who should we be testing now? Without a doubt, all patients with newly diagnosed stage four, many of them can be treated with a single agent if it's targeted. So that's really a population you don't want to miss. But I want to bring up that now the indications are broadening. You can even consider it for squamous cell carcinoma. Sometimes it's a small biopsy and you wonder if you have a concomitant, a mixed pathology where you have squamous and adeno. Maybe it's a non-smoker. You almost never see a non-smoker that has squamous cells. So that brings up the possibility of something abnormal. And then keep in mind that there are a percentage, 5%, maybe up to 10% of squamous cell that may have targetable mutations. So really, it used to be just adeno. And one of the things we're getting away from is that testing where it's algorithmic, where we just do EGFR, ALK, see what those are, then move on to the next option. We're going right ahead and through recommendations from NCCN to use broad-based platforms for testing like next-generation sequencing, which I'll get to. And now, even in resectable cancers, they're doing molecular analysis because the ARDORA trial in the New England Journal, patients even as low as stages 1b all the way to 3a patients treated with adjuvant therapy prior to surgery benefited in terms of three-year disease-free survival if they had a EGFR mutation and received OC-Martinib for three years. Now, it's really, really broadening. It's becoming very mainstream. This is just not for us to go through one by one, but to show that this list of targeted therapies, both FDA-approved and on clinical trials, is ever-expanding. Every time we give this talk, we have to update it. And if we talk about FDA-approved, we're probably at 10 plus one PD-L1, 11 things if you count PD-L1. So, this is just a slide from NCCN saying these are the bare minimum now you should check that have FDA-approved targeted therapy and even consider testing in squamous cell carcinoma, particularly, as we said, in non-smokers, very small specimens or mixed histology. This is for adenocarcinomas. You should note that 50% of adenocarcinomas may have targetable mutations. So that's a pretty large percentage now. And then EGFR remains the most common. There is now a KRAS drug, FDA-approved for KRAS G12C mutation that's KRAS is your next most common mutation. And then there's numerous other driver mutations that are out there that we're going to see clinical trials of forming over the coming years. So this is very opportune. Our moderator, Dr. Fox, was the first author on this paper, congratulations. And I think the impetus for this is it's been shown in the literature that there's really a low percentage of pulmonologists performing EBIS globally that are sending molecular analysis. Some studies have shown even as low as only 20%. So I think that's what piqued his interest and his research group's interest in looking at what factors are associated with ordering or not ordering these molecular tests in patients when we know if they have one of these, they may have improved survival actually. So it's actually very important. So the percentage was not great, but it was still a very large number of pulmonologists, almost 500 responded. And it was a mixed bag of a real world representation, interventional training, advanced diagnostic training, academic practice, private practice. So it was a well-rounded study. So what did they find? Well, they found that folks that seem to focus more on lung cancer, like interventional pulmonologists, academic practice, that they weren't having to master a hundred different diseases. They were a little more focused. Or ones that had higher volume of procedures had an increased compliance with the guidelines for things like recommended passes. Academic pulmonologists would actually either perform or refer for EBIS more often versus just sending for a needle biopsy and then having to restage later. And then the testing rates were also associated with those with interventional training, academic setting and the presence of an institutional policy like a reflex test when you had a non-small cell carcinoma or a tumor border thoracic group that had a policy for it. And then unfortunately it doesn't seem to be spreading out into the community as much as we like. And there were lower rates of testing in community settings and in places with lack of an institutional policy. So this is just a little slide kind of like how we do it. Now there are different ways to do this. This is how most people process their specimens after each pass. A pass is a sample that goes separately to pathology, whereas the needle strokes is a different topic. A pass is a separate sample. And what you want to do is really not put a lot on a smear for a couple of reasons. One, you'll waste a lot of material with clumping and you won't be able to see morphology very well without a thin smear. So that's the first principle. The second principle is most people are collecting material for a cell block, a paraffin-embedded centrifuged pellet that can be sliced and stained for immunostaining, et cetera, or sent for molecular analysis. And the more you put on the slides, the less you have for that. So once we get confirmation, if we have rows of cancer, we start collecting material for cell block. And I'm going to get into how many to do, et cetera. Now you can do these procedures without Rapid Onsite. I'm going to give some rationale for why it can help you with your yield and molecular analysis. It's been shown in the literature not to help with diagnosis. You can get the diagnosis whether you have rows or not, but it can be a real-time feedback that you're getting cellularity that then you can focus in that area to get more cells for a cell block. So there's different reagents that you would instill your specimen for cell block. RPMI, Roswell Park Media, is good for lymphoma. Some places use that. Many places or most use another buffered solution like acetylite where the cells are preserved and it kind of eliminates mucus and some lysing of red cells so you get a more pure sample of the cancer cells. And that's what you prepare your cell block. So really, and that's a picture of a cell block at the bottom, really what you want is a close collaboration with your cytopathologist and you want to meet with them, have a cup of coffee, figure out exactly what their workflow is and how you can feed into it and work together. You can't do this without that. That's probably one of the take-homes here. And along those lines, you don't want to waste a lot of material for immunostains. You really only need a couple squamous markers and a couple adenocarcinoma markers. So squamous differentiation, squamous cells have cytokeratin markers for that, CK5 and 6, also P40 and P63. And then your markers of glandular proliferation, the most common ones are your TTF1 thyroid transcription factor, also present in thyroid cancer, but really lung cancer of the adeno is the other most common one which can help you make that diagnosis and Napsin A. And all these other markers like villin for colon cancer, markers for breast cancer, if that's not in your differential, you shouldn't really be wasting material for that. So that brings us to one of the big breakthroughs, which is next-generation sequencing. When we gave these talks 10 years ago, we were talking about isolated testing for each gene, like a PCR or a FISH test for EGFR or for ALK was FISH. So now next-generation sequencing allows you to test for a whole battery of molecular testing through this parallel DNA sequencing, and you can identify hundreds of genes from just a single test. So it's been a real paradigm change, and it's being recommended by the NCCN as your go-to test to do a more broad molecular testing. Many times this can be done in-house. There's also send-out tests. There's different assays with different cell requirements, but you can really, it's a very powerful tool. And it can be more sensitive in specimens with low tumor cellularity because of the parallel sequencing technique. It can be more cost-effective as one test versus multiple tests in traditional Sanger sequencing and traditional FISH sequencing, doing multiple of those versus one test. And then driver mutations can be detected about half the time in non-small cell, and then about half of those can be treated with targeted therapy. So how has EBIS done? How has EBIS performed for genetic testing? Well, as we said, 10 years ago or so, five, 10 years ago, we were evaluating EBIS only really for a single testing of EGFR, ALK, ROS1, and it showed that it was quite, performed very well. EGFR, 94% had sufficient for that, ALK, 95%, ROS1, 83%. And then there started to be some thought about how do we standardize this? How do we optimize this with the number of passes that we put for cell block? And I will talk about some other, another technique for DNA analysis called micro-dissection where you can scrape DNA off of a slide, but in terms of a cell block, this was a study by Jarmus et al. at Hopkins, and they looked at consecutive cases of non-small cell, but they only tested by a single gene testing PCR and FISH for KRAS, EGFR, and ALK. And they found that a median of four passes was what was the, there was a plateau after that and they got the 95% of those were adequate for genetic testing. But again, this wasn't next-generation sequencing. So what about EBIS for next-generation sequencing? So this is a larger pooled analysis, 21 studies, over 1,000 patients were included. The pooled proportion of adequate samples was 86%. And then they were actually able to, it's quite a bit of, when you consider a nanogram of DNA, nanogram is how you measured, is how DNA is measured, is about 300 cells. So that's quite a bit of cellularity that they were able to get with EBIS for a pooled mean weight of 860 nanograms. Like many of the meta-analysis in this space and that are in this talk, the criticisms are they're very heterogeneous. There are statistical strategies used to mitigate that heterogeneity. So that's where some of these points came in, but they did show in this study that the more you did, obviously, the more you pass as you put in cell block, the better your yield. And it looked like it actually went from four, and then when you got up to six, it was even a little bit better. So this is a bubble plot showing the adequacy on the left, 1.0 is 100% adequate. And then as you go with the number of passes on the x-axis, and you can see when you get to six, you had your highest adequacy for. So it's somewhere around four, we know at least four dedicated is recommended and potentially up to six. So this is a randomized trial of EBIS TBNA with and without rapid onsite evaluation for lung cancer genotyping. It was by Trisolini in Italy, and they looked at patients that had lymph nodes and suspected advanced stage non-small cell lung cancer. And they looked at whether ROSE was used or not to collect molecular material. And in 91% in the ROSE arm and only 80% in just the EBIS without ROSE were sufficient for molecular. So there was a greater success rate with ROSE. There's a lot of literature, as I mentioned, that ROSE doesn't make a difference in diagnosis, and that's true. But we think that with molecular analysis, it can tell you I'm getting a cellular specimen, I'm going to stay here and get more from this lymph node. It can shorten the procedure. In this study, at least, it was more likely to be terminated after a single biopsy site. And the big conclusion was that one out of 10 patients did not need an additional diagnostic test like a media stenoscopy or a repeat biopsy when they used ROSE. So this was a fairly compelling but small study. It was randomized that we go to argue the point that ROSE really is valuable in EBIS, even though it's not necessarily helpful in making a diagnosis. So we talked a little bit about cell block. We talked a little bit about DNA requirements. It really depends on the NGS platform that you're using and the type of technology, as well as the number of genes that the platform studies. So large panels greater than 200 genes may require 50 nanograms. If each nanogram is 300 cells, you can get an idea there. Whereas a smaller panel, like some of the in-house testing, may require only 1 to 10 nanograms of DNA, so somewhere around 300, 500 to 1,000. And then there's a picture down here on the bottom right of the micro dissection technique, where the cytopathologist finds a very concentrated area of tumor cells. And that area can be scraped off and tested for NGS as well. And that's another very effective way when your cell block is not adequate. And then many centers are doing that as a first line. So there are different techniques. And you have to talk to your individual hospital and come up with an algorithm and a flow. So the other big breakthrough in lung cancer, immune checkpoint inhibitor therapy, PD-L1, you can think of as a break on the immune system. And the inhibitor, the immune checkpoint inhibitor, inhibits PD-L1. So the break is turned off and the immune system is revved up. It's not chemotherapy. So when Adam and I were first doing EBES for trials with immunotherapy, they needed a core or a surgical specimen. They did not accept a cytology specimen to make that patient eligible for a trial. That's changed. Now it's acceptable to use cytology specimens for PD-L1. If you look at the top right of the screen, this is from the same tumor. On the left is a surgical specimen. On the right is a cytology specimen. And you can see the PD-L staining in brown. And you can see how it correlates fairly well. And that's what I found was the take-home message from this study. So they looked at consecutive patients with non-small cell undergoing EBES. They used cell blocks for PD-L1 testing. PD-L1 testing is not next-generation sequencing. It's not PCR. You don't do it with that technique. You do it with an immunostain. And they looked at 265 EBES specimens. And they found that about 87% were adequate for PD-L1 testing. And of 34 patients with both histology, a surgical specimen, or biopsy, and a cytology specimen, they had a very high concordance, 91%. They looked at biopsies from the same site. They were also very concordant. So it's showing that it's a pretty well-validated modality. Interestingly, and this has been reproduced in a few other studies, when you look at metastatic sites versus the primary sites, they tend to have a better, a higher PD-L1 score. And the theory being that if you have a cancer that's evading T cells and evading the immune system, they probably have a higher score. So that's the theory there. And that's been shown in other studies. So this was another published study about EBES for PD-L1 testing in non-small cell lung cancer. This was across six centers in the UK and one center in the US. So over 500 specimens. And in the EBES group, which had 189 specimens, the overall percentage of patients with successful PD-L1 was 94%. There were no significant difference in sampling with other methods of acquisition, forceps biopsies, surgical biopsies, et cetera, et cetera. Older patients had a higher failure rate for PD-L1 testing. And then kind of along the lines of what we said in the last trial, advanced end stage, so signs of more metastatic disease, signs of brain metastasis had a higher PD-L1 expression. Again, the thought process being that that tumor is evading your T cells and would have a higher score. So this is getting into a few studies on needle size. This is kind of a controversial topic. The reality is that no one can say one needle size produces a higher yield than another. That has not been reproduced in randomized control trials in comparative effectiveness trials. This is a systematic review and meta-analysis. And it looked at 14 studies, very heterogeneous studies. Some were small case series. Many were retrospective. The overall sensitivity for EBUS was 88%, which is in line with the literature. And then they looked at different needle sizes. And again, many different, multiple different studies were used. The sensitivity was about 92% for 19 gauge. And then the 22 and 21 were fairly similar. And then it seemed that there may have been a trend towards better sampling for single gene testing with a 19 gauge, but it made no difference in next generation sequencing, whether you use 19 gauge, 21 gauge. And the reality with a lot of the studies with needle sizes, you think you get a bigger sample, but what you get is more blood. So it looks like you're getting more, but in reality, you may be getting just as much with a smaller needle. The cells are acquired by the shearing force of the needle. It's less so by capillary action, which draws mostly blood, not so much the cells. So a small, sharp needle can also get you a lot of cells is the theory. And so again, needle gauge didn't seem to matter for next generation sequencing. These were two other studies, one with 85 samples compared the 22 and the 25, and there were no differences in success rates. The second one, 21 and 22, no difference. So what are people doing to get this precious material? What are some other things on the horizon that we can look at to improve our yield? This is a meta-analysis of intranodal forceps biopsies versus standard TBNA alone. For adenopathy, again, small number of studies, six observational studies, only about 400 patients. And then the overall diagnostic yield was very low for EBUS TBNA, which seems out of line with what's been published, but it was quite high for the combination of forceps biopsies with standard EBUS. There really, there was a trend towards more complications. As you can imagine, some of these lymph nodes have vessels in them, but they were not out of line with what you'd expect in a transbronchial biopsy or excessive to say that the procedure seemed too dangerous. And they were lower than they would be with transbronchial biopsies. So just something to ponder, something to think about, particularly when you're not getting what you need from your needle samples, and you perhaps are not able to do liquid biopsy or other ways, and your kind of back is against the wall, it's something to try. This is the technique. In the study, they used a 19-gauge needle. This is a review on the technique by Chang et al, and they make a needle hole with a 19-gauge needle. This is the tiny forceps biopsy that you can see under ultrasound, go through the hole and take a sample from the lymph node. In the top here, you can see just what a 19-gauge needle sample looks like, a lot of blood, and intranodal forceps, it looks like you get more tissue, a lot of blue, meaning a lot of either pathology, lymphocytes, et cetera. So at least something to ponder and to think about. And that brings up another publication recently, looking at a similar idea, but with a cryoprobe. And this was a study done by really, really world-class expert bronchoscopists, Felix Hirth and some other pioneering groups in China. So it was three groups in China and a group in Germany. And these are people that invented this technique. And actually they looked at, it was a randomized multicenter trial, open label, and they looked at four passes with a needle and they made a small incision with a high-frequency or electrocautery needle probe. And then through that, they went in and got one cryobiopsy. So that was considered the combined technique, which was four needles, one cryo versus four cryos. And if you look at a couple of their graphs, there really was not, interestingly, a difference in complication rate of statistical significance. You might say, you know, there may have been a little more bleeding. None of it was statistically significant. Overall diagnosis, no definitive diagnosis was less in the group for combined with cryoprobe. However, when you look by disease state, like when you look at lung cancer, there's really no difference. It did not help diagnostic yield. It did not help diagnostic yield in lymphoma and other diseases, although small numbers. So that was traditionally the area that we were also very interested in and it did not really make a difference. It did seem to diagnose benign disorders. So you could get adequacy for lymphocytes where you may be just getting blood in another one. You could get sarcoidosis, granulomatous disease. So there was a trend towards that, but in the area for this talk that we're looking at in lung cancer, it didn't make a difference. So what about genomic testing? So that's where it seemed to be a little bit better. Keeping in mind, the study was well-powered for diagnostic yield, but not for, look, this was a secondary endpoint. So lung cancer genomic testing and PD-L1 were suitable in almost all of the samples, 97% when they used this combined approach of needle followed by one cryobiopsy. And they had a relatively low number of cases where genomic testing was available, 79% when it used the EVIS alone. So again, something to think about, but further evaluation will be needed. So what are some of the challenges that we run into? Why do we struggle sometimes? Why are we not 100%? And what about the patient population makes tissue-based testing such a challenge? Well, it's invasive. So if you look at this, you can see the needle coming down into this mass with a 15% pneumothorax rate, a 6% bleeding rate, et cetera. Many of these patients have to be hospitalized, not something you wanna do with an advanced stage cancer, a patient that's, you also wanna factor in quality of life. You know, we hate it when we have to do a repeat biopsy. That's a potential challenge. And then, you know, some of it is operator. You know, people that have done thousands of EVIS versus people that have done 40, et cetera. You know, it's not all numbers-based, but there's some feel, there's some amount of speed that you may have that you may cause less bleeding, et cetera, in your sample. So there's operator factors, there's bleeding that can make your sample difficult to interpret. The big one is necrotic tumor. I mentioned that big liver lesion in the early slide. It looks like a tempting area to biopsy. A lot of times you get dead cells and there's not DNA for molecular testing, unfortunately. And then there's another topic called tumor heterogeneity. And that may explain why we're not successful with tissue sampling for molecular analysis. So you can have it as intratumor, where in B, you see in this tumor, there's four different types of molecular profiles. So depending on where you sample that lesion, you will get a different answer. There's also intermetastatic, where say the brain met harbors this mutation, the liver met may have a different mutation or lack of mutation. So they're not homogenous in each tumor site. And then also in a single metastatic focus may also have a different molecular profile. So these are just some of the challenges that we're dealing with. So to kind of summarize a third of the patients, these are additional challenges. A third of the patients with advanced non-small cell may die within the first two months after initial diagnosis. So just think, if your biopsy takes 10 to 14 days to come back for molecular analysis, if you have to do a repeat biopsy, if you have a busy program and it takes a couple of weeks to get the patient in, you may be approaching that two months and you haven't even gotten treatment started, unfortunately. Now that may fall outside of NCI guidelines, but in the real world, that does happen. Poor ECOG status of many patients, they don't wanna go through invasive procedures. More and more we're needing shorter turnaround time. And then up to 80% of patients with non-small cell lung cancer with advanced disease will only have tissue from small biopsy sites. So they're not getting large amounts of material. And then up to 31% don't have accessible tissue, up to 20% are inadequate due to insufficient amount. So that gives you some summarizes some of the numbers that are out there. And this is kind of a classic case that puts all those things together. This is a 47 year old woman, a heavy smoker, has a history of rheumatoid arthritis, and she presents with respiratory insufficiency hypoxemia. And you can look at her CT and you're already very suspicious of cancer. You see interlobular septal thickening, you see some small pleural effusions, you see a big Hylar mass, adenopathy, multistation. So you already think, oh, this ought to be pretty easy to diagnose. So you can do an E-bis, you may try the pleural effusion as we talked about, but it was very small. It's not all malignant effusions are always positive. So ended up with this patient doing an E-bis. It was positive for several immuno stains, but mainly in the necrotic region. It was read as there were scant viable tumor cells and massive amounts of necrosis. So they could not make a definitive diagnosis, but it was suspicious for adeno with squamous differentiation, and they recommended further tissue confirmation. So you can imagine this person that's hypoxemia and sick, you know, what can we do? So a potential solution for such a case that brings us to one of our final points, and that's circulating tumor DNA. Cell-free DNA isn't all tumor DNA, but when you say circulating tumor, that's the DNA just from the tumor cells. It can be found in plasma. And this was a review article back in 2018, and it kind of predicted the great surge in this technique that's happening in the oncology world. So you can get circulating DNA, tumor DNA from several, from three main sources. The tumor can have apoptosis, so it's releasing cell-free DNA. There is DNA and exosomes from tumor cells, and then the tumor cells excel themselves. So that can be extracted in the plasma, and then they can run next generation sequencing on that. So this patient happened to have a liquid biopsy at the time of the EBIS bronchoscopy, and they did actually have a mutation. Now, it was only available, the treatment was only available through a clinical trial for this particular BRAF mutation, but they did have some response to it, and it just goes to show the potential utility for the liquid biopsy among other things. So what are the advantages and disadvantages of it? So if you look at the flow chart here, we've kind of already talked about the top flow of getting your sample from EBIS, making a cell block or getting DNA from smears. You can do next generation sequencing, you can do IHC on the cell block. And that's one of the big advantages is that you can identify histology, which you can't from the liquid biopsy. So you can get pathologic assessment, is it large cell neuroendocrine? Is it adeno? Is it squamous, et cetera? You can assess non-DNA biomarkers, you can assess for PD-L1, which liquid biopsy does not. Again, that's very important. Some patients with high PD-L1 scores, for example, greater than 50% in stage four can just be treated with immune checkpoint inhibitors. So that's very important to find. Disadvantages, it's a longer turnaround time, limited tissue quantities for the reasons we mentioned, the challenges we mentioned, it's invasive. It's also inconvenient on everybody, the patient most to undergo re-biopsies, especially if they're getting sicker at the time of recurrence. And we talked about the concerns about tumor heterogeneity. So liquid biopsy, we talked about getting cell-free DNA from plasma and doing tumor genotyping directly from that. So you can see there's fewer steps, it's a shorter turnaround time. It has a high concordance rate that's been shown in the literature with tissue. It's minimally invasive, obviously. It's reproducible over time, whereas a procedure you may have one that's more effective than another due to various procedural concerns, complications, et cetera. And you're sampling a systemic population of DNA. You're not sampling DNA directly from one site. Typically, you should work around some of that heterogeneity issue because you're capturing the DNA from all the tumor that's being released into circulation. Obviously, disadvantages, non-DNA biomarkers are not available. It can increase costs. It may not be approved in some states, some insurance plans if you're doing tissue testing at the same time, and there are false negatives. So currently, the NCCN advocates testing of circulating tumor DNA for monitoring purposes, particularly when a patient can't medically handle an invasive procedure. It also is useful, as we said, at the time of recurrence, but tissue sampling kind of remains the gold standard if it can be done. No insurance is provided at the start of 2016, but now almost 50% or more of insurances are covering it. Policies are also increasing to cover multiple different types of cancer cell types, and then Medicare coverage policies are improving in this regard. So potential clinical applications. One very important thing when patients on EGFR tests with tyrosine kinases have recurrence, it's very common for them to develop the T790M resistance mutation, and that can be detected on a liquid sample. Potentially using it as a tumor marker. Does it go down with treatment? These are potential areas for more research. Identifying candidates for adjuvant therapy after surgery. Theoretically, with a complete resection after a certain period of time, greater than 24, 48 hours, you may have loss of your circulating tumor DNA, or you should, and if you're not getting that, does that imply that that might be somebody who would be more likely to benefit? And then eventually, would it be a screening modality? This is just a little bit about the T790M mutations. About 50% of non-small, so it's very important, about 50% of non-small cell patients with the EGFR mutations will develop resistance with this mutation, and osimertinib is the standard of care for those patients with that mutation. And so this is one area where the circulating tumor DNA can be helpful, particularly in patients that are medically unfit for further biopsies, et cetera, or from a quality of life perspective, just having a blood draw. So in summary, EBIS is recognized, as we know, as a first-line test for diagnostic and staging purposes in advanced lung cancers and for staging cases prior to surgery, but now it is a validated method also for testing for mutation analysis and for PD-L1. As we saw in the Adam Fox paper, there's an opportunity to improve awareness of best methods for biomarker testing in EBIS bronchoscopy. Paramount is that you collaborate with your local cytopathology department. That's why it's hard to provide why the American College of Pathology and the joint guidelines have a hard time providing a very succinct, exact guideline for how to process tissue because each lab and how they do it is so different. So you really need to meet locally with your group and streamline it and figure out ways not to waste tissue, not to do extra testing that is not clinically relevant. And then we talked about some of the challenges presented to us in a clinical setting, frailty of some of our patients, tumor heterogeneity, a need for repeat biopsies that prompt us to think of another alternative like liquid biopsy. And this will need further investigation and the exact role will be clarified over the next decade or so. So I think Adam has been collecting questions from the chat, so I'll stop there. Absolutely, I'll encourage everyone to put questions in the chat or in the Q&A section. And thank you so much, Dr. Pastis, for a really comprehensive look at tissue acquisition for non-small cell lung cancer. I think you really highlighted the role that pulmonologists can and are playing in initial diagnosis staging and in biomarker testing. And I think we just have our first question here related to obtaining PET-CTs prior to invasive mediastinal staging with EBIS related to the timing. Is there a preference for obtaining PET-CT before or after bronchoscopy with EBIS? Yeah, I think there's maybe two questions there. The first one being the role of it before EBIS. PET scan can help you because it may find sites of disease in the liver, in the bone, et cetera, and help you with staging. So it's useful for that reason. Technically speaking, when you do an EBIS, you should be doing it in a systematic way where you start at a contralateral N3 node and work your way back towards the primary regardless of the PET scan. But the PET scan can still be helpful there. If you don't see anything in the contralateral hilum and you have a PET scan and it's PET negative, you might start right at the contralateral N3. If you have a right upper lobe tumor, you might start at 4L versus going all the way to the hilum and doing every lymph node. So it can help you a little bit with your EBIS, but the primary benefit is that it can show you sites of distant disease in bone, liver, et cetera. Now, the second piece to that is, if you poke those lymph nodes, is it gonna give you a false positive on a PET scan? And I think there has been a study of that. And contrary to what you might think, in the majority of those, it doesn't affect that. Have we seen it? Yes, we've seen some false positives, but in general, that's not a problem. So you wanna get the patient in. Remember our first cartoon that showed timeliness. So sometimes PET scans can take three weeks. So if you can get it in a timely way, it's helpful, but you would not necessarily wanna delay your EBIS by three, four weeks so you can get the PET scan. Absolutely, thank you. We have another question about any recommendations for repeating bronchoscopy with EBIS if and when additional tissue is needed. Are you aware of any data or recommendations on a repeated bronchoscopy? In general, if you're an experienced EBIS center, you most likely, and you're doing it in a systematic way, you're gonna get the same thing. So the idea that I just sit there and do more and more and more is gonna get, there's something about that lymph node that there's heterogeneity in the intranode, there's necrosis, and the majority of those, you're gonna get the same thing. That being said, there have been redo EBISes where we've gotten enough. So in general, a repeat EBIS is not helpful. You may consider another type of biopsy. You may consider mediastinoscopy in the right setting, and you may also consider liquid biopsy. So if you're a very experienced operator and you've done a thorough job of sampling the target, the node that's cancerous, and you're getting necrosis and whatnot, you may not get it on the second. So we kind of had a policy where we would avoid repeating, but there are cases where you're so concerned and you're worried your sample wasn't as extensive as it could be, you can try, but the majority, no. And I think there's a few reasons to consider sampling multiple highly suspicious sites in EBIS, one, maybe to confirm multistation disease in the right setting, but then also if, for instance, your rapid onsite evaluation is showing necrosis, is there any recommendation or practice that you're aware of for intentionally sampling multiple positive sites? Yeah, if you're seeing necrotic material, that would be one. And I think when the new staging guidelines come out, there's gonna be an increased stage for multistation N2 disease. So that's gonna be a little more important for us not to always just stop at the first positive node. So I think we're evolving and to Adam's point, I think we're moving towards probably multistation even when you get a positive. And you did bring up several times in your talk about earlier stage disease. Are you finding that surgeons and oncologists, the multidisciplinary team are interested in having those biomarker testing results prior to surgery? Yeah, I think it can help them plan in those cases with an EGFR mutation. I know there's also an interest in a second primary and looking at the molecular profile of an early stage tumor when it could be a second primary versus a met. So the oncologists are using it not so much for therapeutic, but for diagnostic purposes. So it's kind of like when you say to the pathologist, would you like more tissue? They're never gonna say no. If you say to the oncologist, would you like a molecular analysis? They're never gonna say no is my take on it. I think that's fair. One question about whether we can justify or put into context how different needle sizes didn't show big differences in yield, but that the mini forceps biopsies seemed to make more of a difference in yield. Is there any way to put that into context or contextualize that difference that's seen? It's hard to say in that there are different studies that took different numbers of passes that different operators that did a different number of cases. So a very heterogeneous population. Theoretically, you're getting more architecture with the forceps biopsy and a bigger piece. But then again, in lung cancer, the diagnostic yield is so high, it's hard to make a difference in 90 some percent diagnostic yield. So where we really would like to see it is, can it get us more molecular? That's where we would like that 10, 20% bump in adequacy. And I think you'll know your answer to this one. This was a follow-up question to repeating EBIS. If there's a reason to think that a repeat EBIS would be beneficial, how long to wait to repeat that EBIS? If you're gonna do it, there's generally not a reason to wait any set amount of time. It's more for logistical reasons. You wanna try to get the diagnosis with the timeframe in mind that your molecular may take 10 to 14 days to all return, depending whether it's in-house or send out. So the sooner the better. Absolutely agree on that one. I've got a couple of speculations. Do you have reasons for why rows may or may not really make a difference in some cases or why, or do you think that it really in the end will make a difference in molecular testing? Yeah. Why would you support that? I think it makes a difference in molecular testing. It's been pretty well established that it doesn't improve your yield, your technique, whether you have rows or not, and you're gonna get adequate material to make a diagnosis of non-small cell. But what rows can do is give you some feedback into your cellularity. And if you need a couple of hundred cells for IHC, you need a thousand cells for your NGS platform, and you're only getting one or two malignant appearing cells, that's useful information, but you're still gonna, those one or two cells, when you put them together with multiple other one or two cells, you may still get the diagnosis, but you're not gonna have enough for molecular. So it gives you that feedback. It may also, to your point about multiple stations, if you're getting necrosis, you may, you're getting dead cells, and I'm sure I'm in this lymph node, but what I see under the slide is mostly dead cells. I'll move on to another lymph node. There's a sort of a coagulation necrosis sign. It's one of the few really predictive markers of a malignant lymph node on ultrasound. And if you see that, you may pause and say, maybe I wanna also do another node that doesn't have that, because that usually means there's some necrosis. And is there any best practices for how many sites and how many passes for molecular, for those who don't have access to rows, as we know it's not uniform and it can be very costly to institutions to implement? Yeah, I think getting, you know, the smears can be helpful because if you're dealing with squamous cell, one of those you're putting, if you don't have rows, you're gonna put one of the smears in alcohol. And when that goes to the lab, they do a pap on that. And that can be very good for identifying squamous cell. So you do wanna get a couple smears. And then if you're by yourself without rows, four to six, at least for, leaning now towards six based on a meta-analysis, but the traditional answer was four, but now maybe even up to six dedicated for a cell block in whatever your department uses for that, whether it's cytolite, RPMI, et cetera. And would you, is that, would that be recommended for all sites that are sampled, that number of passes or only for maybe the most suspicious or? It would be if you don't have rows because if you have rows and you see lymphocytes, you only need three passes, you know? So without rows, you would probably do that number of passes for each site that we discussed. And then we had a question asking about- And just backing up on that, when we're staging, we're talking about, you know, over five millimeters we're sampling every station. Absolutely. We have another question. Did you, I'm looking back at your slides now, did you present any of the data on cryobiopsy for mediastinal lymph node sampling with EBIS? There was a question about where adequacy may stand for that. The bigger study that was randomized and multicenter was the Felix Hurst study in the Lancet Respiratory. And it showed that the cryobiopsy of the lymph node did not make a difference in diagnostic yield for lung cancer, but it did help with molecular analysis. But it wasn't powered for studying that. That was a secondary endpoint, but it showed that it could be a good adjunct. Keeping in mind, these are the top, you know, some of the very top bronchoscopists in the world, you know, that are doing this. So whether it's a mainstream test is not panned out yet. And you and I were talking just before the webinar about how each of these studies may or may not include both PD-L1 and genomic testing, because they are slightly different. And that one is one that included both genetic testing and PD-L1. Right. Absolutely. I think, you know, we're running up right to seven o'clock and seeing no further questions from the audience. I'd really like to thank you, Dr. Pastis, for your time and really an excellent overview of procedural and sampling considerations related to biomarker testing in non-small cell lung cancer. I'd also like to thank the organizers at CHEST and the audience members for their attention and their questions. I'll remind everyone that the third and final part of the webinar series is scheduled shortly on June 16th. And Dr. Pastis, I'll let you mention that slide there. Yeah, thank you. I just wanna point out that the Bronchoscopy Domain Taskforce at CHEST, I've had the privilege of working with them for a number of years, and they're putting on a great new course this summer that will go over some of the topics that we talked about today. So for those of you interested in EBIS and maybe refining somebody already good at EBIS that wants to refine themselves a little more, it would be very useful. So I just wanted to provide that little plug. Absolutely. Well, thanks again, everyone. I hope you all have a great evening. Thank you.

endobronchial ultrasound

EBUS

transbronchial needle aspiration

TBNA

non-small cell lung cancer

biomarker testing

liquid biopsy

mutation analysis

PD-L1 testing

tissue collection