Case-based presentations, mediastinal EUS, fiducial placement, elastography, and contrast-enhanced EUS. It's a bunch of topics. I'll try to go with the high-level points on these, especially in the interest of time, but also because some of these are not being practiced everywhere at this point in time. Some are coming along, some are in and out, the frequency can vary, and so forth. So we'll review this, and then we'll have a discussion. We want to save some time for questions as well. My disclosures, as you've seen before, the objectives of this talk are to speak to the evaluation of mediastinal pathology, mediastinal EUS was very, very popular in the late 90s and early 2000s until the advent of E-Bus, but it still has a role, and we'll talk a little bit about that. The value of an EUS-guided fiducial placements, I will review that a little bit, and the technique, and so forth, and where it stands today, and finally wrap up with contrast-enhanced EUS and EUS elastography, and what they might have to add in our diagnostic evaluations. The format that I've followed here is presented here. I'll try to present the concept first, maybe one or two or more cases, and then the technique, a little bit on the data, and then some takeaways. So starting with mediastinal EUS, mediastinal EUS really still remains one of the common indications, at least in our practice. The concept here is to evaluate the mediastinum and the periesophageal structures through a transesophageal route. The mediastinal targets should ideally be contiguous with the esophageal wall. They may not be contiguous as long as you don't breach the pleura, and the pleura is a very bright line, almost akin to the wall of the gallbladder, if you can imagine, that exists in the mediastinal evaluation. You do not want the needle to cross that, because more than likely, you will have a pneumothorax. Mediastinal adenopathy, typically the case that we'll get will be a metastatic malignant rule-out, or it could be a primary malignancy, such as a lymphoma. Contiguous etiologies are less common in the Western world, but are a little bit more common in the Far East. Esophageal cancer staging, certainly for staging and fiducial placement, as I will show you soon is an important indication as well. Centers like ours, we do get often called in to evaluate lung masses. These are proper lung lesions that are now contiguous with the esophageal wall, and they are up for grabs in terms of safe FNA or FNB. We are a center where we do about 600 e-buses a year between thoracic surgery and interventional pulmonology, but still there is an element of non-diagnostic sampling, or stations that are more easily amenable to us that get sent over. And then finally, this has been talked about and is one of my favorite topics to talk about is cyst evaluation in the mediastinum, and I fully endorse the cautionary notes provided by the other two masters in this regard, which is to say that most of these are either duplication cysts in the esophagus or benign mediastinal cysts that should not be intervened, at least through an endoscopic route. So stay away from FNA in these lesions. A basic evaluation of mediastinal lymph node stations, and the pink organ is the esophagus, gray is the trachea, and then you have the vascular structures. I have highlighted for ease of reference those lymph node stations that are typically in the realm of EOS. These are station 4L and 4R, the so-called aortopulmonary window stations, the very easily seen and accessed subclarineal station, and of course the lower esophageal lymph nodes in the inferior mediastinum. These are the easy ones for us, and probably the most impactful one of these is the aortopulmonary window lymph node, which is very difficult to access sometimes with EBUS and almost very difficult to access even with thoracotomy. So if you ever get an aortopulmonary lymph node consult, please take it because you are probably the only person who can sample that. This is a case actually from last week. This is a patient who was referred to me, a 53-year-old smoker, history of alcohol use, has some dysphagia, underwent an upper endoscopy in the community, and was found to have a lesion in the esophagus. These are my endoscopic pictures. And he was referred with a negative CT and PET. Now these are both items that Dr. Mishra alluded to. Completely negative staging workup, sent to me for EMR. Now if you look at this endoscopic evaluation, I almost always use a cap for forward viewing endoscopy. There's a bulky lesion here at the GE junction, there is a stricture on this side, and then there is a valley here. This is where the tumor is infiltrating. And then there's another bulk that extends into the cardiac. This is not a lesion you want to EMR. This is perfectly well suited for an EUS. Narrowband imaging per protocol is always done. I don't think it adds too much. In this instance, the luminal endoscopy with high definition white light actually reveals that this is likely an invasive lesion. So here is the EUS T staging of this GE junction carcinoma. And you can see here, over here the lesion starts relatively superficial. You can identify the submucosal layer, but first the submucosal layer goes, and then at the point of maximal infiltration, the muscular dyspropria is invaded. So the endoluminal findings are correlating finally with the EUS findings, and at this point the T stage is that of T2, and we are now withdrawing all our preparatory things for EMR. And now moving towards further staging of this tumor. Further staging of this tumor reveals a very discrete, dark, hypoechoic, round lymph node in the vicinity of the tumor. The tumor is right up here, and this is not a lymph node you want to sample because you're going to seed it if it's benign, but most likely this is a malignant lymph node. Now further staging of this patient and the same session reveals that he has a auto pulmonary lymph node, which is right here, it's about 10 to 12 millimeters in long axis. This is the aorta, this is the pulmonary artery, and this is the needle path. And at this point you have taken a G-junction tumor that was referred for EUS, and now has become a widely metastatic lesion. This is exactly the video from last week in preparation for this course. And you can see here the needle coming into a transesophageal route. Now I'll pause it here for a second. You can see that this is where we use a door knock technique, which I referred to in my first lecture, which is the needle comes in, you're barely about four or five millimeters of traversal length available. You have the pulmonary artery staring you right in the face, so you have very limited ability to do anything in terms of errors. So we try to use a door knock technique, and you can see that there is some fibrosis in the lesion over here. And initially the needle is not traversing like the way we want it to traverse. And I deliberately left that in the video, so that when you can see that when ideally the FNA is being performed, the needle should traverse through the lesion, still having difficulty until we go with the door knock technique, and now the needle is moving fully, and now you'll get a better sample. This of course was positive for malignancy, and this patient was definitely referred on to that. So in comparison, this patient had had an EUS at the outside center, and was found to have a pathologic quote unquote pathologic lymph node in the subcarina. This is typically how a lymph node will look in the subcarina, so please note the normal anatomy of the subcarinal lymph node. These are not lymph nodes that have to be sampled. This is an elongated dumbbell shaped lymph node, which is relatively, you know, isoechoic or more gray, so to speak. It's not as dark, it's certainly not round, and this is not to be sampled and does not count towards the N1 staging of these lesions. Now this is a paper we wrote not too far back with my good friend Sham Thakkar, who's now in West Virginia, which really summarized the entire world literature on EUS guided staging for esophageal cancer. All the references are here for those who are writing on this topic going forward. The TNM staging is listed here, and you can see here when the EUS imaging of esophageal cancer in, you know, the mediastinal stations through the esophagus is sharp, it can be very sharp. You can see all the layers, even with a dedicated echo endoscope, and these are typically the imaging findings that you'll see with a mini probe, but sometimes you get lucky that you can get that type of imaging. So another case here, lung mass with mediastinal lymph nodes. This is a right lower lobe lung lesion. This is a hyaluronid lymphadenopathy, patient presented to pulmonary and they referred on for EUS. This is a pathologic lymph node in the subcarina, very different than that that is, that I showed you as a normal variant. So recognize the normal variant versus the pathology. And again, these are relatively easy to sample and get the diagnosis from. A brief comment on, these are again old studies because remember mediastinal EUS kind of peaked in the early 2000s, so this is 2002 by my mentor and teacher Mark Catalano and others showing that the diagnostic rate was very high, even in the old days, almost two decades ago for mediastinal adenopathy for a malignant process. Now sometimes you get doozies. This is a patient, very, very large subcarina lymph node, in fact, two thirds of the mediastinal circumference of around the esophagus is consumed by this lymph node. This patient had a remote history of prostate cancer. And lo and behold, you do an EUS and you find prostate cancer with the PSA antigen standing positive. So sometimes the etiology is not routine, go with an open mind, and you'll find sometimes very strange diagnosis at play. Now in the non-Western population, tuberculosis and sarcoid tend to be a very high on the differential diagnosis. This is another older study by Nat Fisher-Ravens, another pioneer in EUS. Back in the day, 52 patients and accuracy for both tuberculosis and sarcoid was very high in the mediastinum, provided the appropriate cultures and methodology was used for diagnostic stains and studies. So the overall data on mediastinal EUS looks pretty good. These are also reflected in the ASGE guidelines, which you can easily refer to. So again, this is for malignant processes more than for lymphomas, which I alluded to earlier can be a little bit harder to diagnose because of the need for a significant tissue that needs to be available. So the pearls for mediastinum, one thing that you will learn in your fellowships and your early practice is that the radial versus linear anatomy of the mediastinum can be very different compared to the abdominal imaging. I feel EBUS and EUS are complementary and still have a role in the, what we call the complete medical mediastinoscopy paradigm. EUS should be incorporated in patients when the clinical suspicion is high and that it'll alter the management. FNA in the mediastinum is safe as long as the pleura is not bleached, breached. And in general, we will avoid the FNA of cystic lesions in the mediastinum and keep an open mind because all kinds of diagnostic outcomes can be present when you are sampling nodes in the mediastinum. So moving on to fiducial placement, the concept here is again, transluminal placement of solid gold markers. That's about $150 a pop using the FNA technique. You can use floral assistance in the beginning just to get the three-dimensional spatial orientation right. But as time goes on, you can do these purely EUS guided, which is fine. And the goal is to minimize the radiation collateral. That's the main goal is to target and pinpoint the treatment. And for this, you do need specialized radiotherapy equipment such as that is listed here. Pretty much right now, we have the preloaded needles available, the reloadable needles are probably still available, but it's very cumbersome to load them. So the locations are up for grabs. I think you can put it in most any solid organ that you can access with EUS and the technical success rates are very high, although migration, infection, chest pain, and so forth can occur in a minority of cases. Now the two basic commercially available needles are listed here. This is one particular brand, which has four preloaded needles. These tend to be a little shorter in their overall length. And then there's another brand that has only two preloaded needles. And these are also variable size, but each of these sizes is longer than the one over here and either one can be used and they're available in different gauges. So the delivery of EUS needles preloaded versus reloaded, obviously the preloaded lesion needles have a shorter procedure time. And that makes sense. And therefore, for the most part, most folks are doing fiducial placements now are using the preloaded needles. This is a fluoroscopic image of how the FNA is performed. And as you push out the fiducial, you withdraw the needle back. And then one by one, I can just show that again, you push out the fiducial and you withdraw the needle, sorry. And then you can see it, I'll just let it play. So there you go. And then they have to be deployed in a spatially disoriented manner so that they provide the maximum target for the simulation. And here's the EUS correlate where the fiducial is coming out. And that's how it is typically done. But as I mentioned, you can do it just with EUS guidance. So case studies, this is a difference between a SBRT simulation where the fiducials are placed in this lesion here, as opposed to the standard radiation simulation. So you can see the amount of collateral that is reduced. And that's really the name of the game in this particular intervention. Here's a slide. I think this was a paper or a chapter we did some time back where we highlighted the demo of how it looks in the pancreatic mass, how the fiducials are spatially disoriented, and how they appear when the procedure is finally completed. This is the CT correlate. And this is the fluoroscopic correlate. This is a case where a biliary SEMS was placed in a patient with a pancreatic malignancy and was causing all kinds of scatter. And the patient has kind of maxed out their radiation dosage. And at this point, we were asked to place a fiducial. This is the scope position, which is ideal for pancreatic fiducial placement. Because remember, you are driving markers through. So a straight scope position is important if you can achieve that. And then once we place the spatially disoriented fiducials, they were able to better target the active tumor environment and not have to deal with the scatter from the SEMS. This is a different patient, an unfortunate young lady who had upper rectal tumor nearly obstructing. And the ask was to place fiducials and also provide palliative decompression as a bridge to surgery. So this patient, at this point, we placed an endoclip at the distal end. We used to do that in the beginning for colonic lesions because EOS can be a little tricky getting up a little bit into the upper rectum or early sigmoid. And then the under fluoroscopy, FNA into the mass, and then spatially disoriented lesions to facilitate rectal radiotherapy. This is an abstract I think we had in 2019, looking at about 11 patients in the conventional radiotherapy arm and 11 patients in the SBRT arm. This hasn't made it to a paper yet, but that has reminded me I'll chase the co-authors to do that. But the conclusions from this 22 patient study was that the use of fiducial markers in our cohort allowed for more accurate delivery of radiation, a more complete response to treatment, and also was considered for esophageal cancer patients. This was important for facilitating a more targeted radiotherapy in this population. So there's an early abstract there that we did internally and was presented at EDW. This is kind of the overall data on EOS fiducials. This is 500 patients, retrospective review, very high technical success, relatively low adverse event rate. And the pancreas, again, another 350 patients, low AE rate, high success rate, rectal cancer, esophageal cancer. These will be all part of the enduring material. But again, the message here is that the data on outcomes from the fiducial is relatively less. The data on the technical success, the feasibility of doing it, and so forth is a little bit on the high side. This again is something that I would like to see practice more and more some of the outcomes data being published, but certainly an easy technique to learn and a service that we can provide in a multidisciplinary fashion. So the pearls for EOS fiducials are listed here. I think the big messages here are preloaded needles are pretty much the mainstay at this point. Once you get some experience, a fluoroscopy is not necessary for every case. Antibiotics have been used traditionally. Again, a great study on antibiotic use has not yet been done in my mind. Ideally, these are done at the same session as FNA and FNB. We do not know what the ideal geometry is, but three fiducials that are spatially disoriented are the recommendation from our radiation colleagues. I think the length of fiducials can be an issue for small tumors in that they may either go into the mesorectum if you're doing rectal fiducials or in the mediastinum if you're doing esophageal or they may more likely fall into the end on the endoluminal side. But whenever you are contemplating fiducials, introduce them to your tumor board and make sure that you have the specialized radiation treatment platform before you embark on this. Final section here is contrast and elastography. I'll try to breeze through it a little bit in the interest of time, so we can get to some questions. So for the vast majority of uses, contrast EOS and elastography are additional imaging platforms over and above the EOS platform in the same machine at the same session for differentiating these malignant pancreatic lesions that are listed here from one from the other and also non-malignant entities sometimes, especially autoimmune pancreatitis and certain types of serous cyst adenomas that behave differently. Now contrast enhanced EOS again focuses on improved visualization of blood vessels in the lesion. I won't call it tumor because it's not always a tumor and the identification and behavior of the parenchymal vasculature is where the money is. And you can see here that some of the more recent studies have shown that the sensitivity and specificity is pretty high, certainly higher than it was in the earlier days. And it's certainly as I think Dr. Nantali had shown that post ablation monitoring is becoming a very important role or key indication for this technique. This is kind of sort of how the tumors or lesions will behave. So if you have a hypovascular hypo enhancement pattern, that's most likely an adenocarcinoma. If it's an isovascular pattern or iso enhancement, it's inflammatory. It also could be autoimmune pancreatitis. If you have a hypervascular hyper enhancement imaging, then it could be a pancreatic net. There are nuances to it, which I will not be able to get into it, but at least let's look at a few case studies. This is what I call an isoechoic pancreatic mass. This was a patient that had been around with several EOS's, CT scans and such. And it wasn't clear whether he has any of these diagnoses that are listed here. So contrast enhanced EOS suggested a uniform isovascular kind of enhancement and putting together the serological markers, the clinical picture, the test of time, the EOS imaging, contrast enhanced EOS imaging, tumor board, the final diagnosis was autoimmune pancreatitis. Patient was treated with appropriate therapy and over time the EOS imaging changed. So on that note, a nice paper here looking at contrast enhanced EOS for AIP versus cancer and found to be useful in this setting. So for those folks who are involved with considering enhanced imaging, this may be an area where we could put it to work. And certainly a lot of work on this has been done in the European continent and hopefully more to come on this side. Now a classic imaging for pancreatic cancer is depicted here. This is obviously an ugly looking lesion in the head of the pancreas. And just for proof of point principle to show that the hypo enhancement or hypovascularity of a malignant tumor compared to that, what I showed it for a benign condition. So it can be really remarkable, but like I think it was Dr. Amadou said, the extremes of imaging are really very helpful. It's difficult to pass out things when things are a little bit in the intermediate zone. So I'll skip these ones in the interest of time and just show you another example of a benign lesion. This is a serious microcystic cyst adenoma and there's a macrocystic component. And you will see with contrast enhanced EOS, the microvascular enhancement in the septae, the septae are the more vascular areas is very, very sharp. So just a proof of principle. For elastography, the concept here is measuring tissue stiffness and hardness used to be hue histograms. Now we use strain ratios. And the newer thing on the block is shear wave elastography, which I will not be able to get into, but safe to say the main purpose of elastography is to differentiate soft tissue from hard tissue. And cancer tissue is typically hard tissue. And we measure that by measuring strain ratios. So cancer tissue, hard tissue is blue on elastography and soft tissue, benign tissue is more towards the red spectrum. So here is an example of a pancreatic lesion. Again, multiple FNAs, some solid cystic areas, and then EOS elastography showing where the needle should go, where the more deep blue areas are, and it may facilitate a more accurate FNA, FNB protocol. Same thing here has been invoked for smaller lesions, lesions that are less than two centimeters. Certainly those with cystic areas, which are difficult to get a yield on, EOS elastography tell you where to go with that. And nowadays with the more latest modules on the machines, this software is available in the older days, elastography was not easily available here to purchase it separately and was quite expensive. So if you have interest in it and time and applicability in your clinical practice, this may be a paradigm you want to invest in. This study I want to show just primarily published in Endoscopy to say that whenever you have a soft target identified by elastography, the chances of a malignant diagnosis, particularly in adenocarcinoma are extremely low. So the negative predictive value of elastography is something we hang our hat on, but the specificity in other areas can sometimes be variable, which is part of the reason why it has not caught on widely. So this is another study here, relatively more recent, looking at how do we perform when we combine EOS elastography and contrast-enhanced EOS. So you can see here the numbers are in the 80s and the combined is about 93, 94%. So approaching the diagnostic yield levels of FNA, but obviously in the Western world, tissue still remains an issue and that's where some of the challenges are. I mentioned shear wave elastography. This has been available in the hepatology realm for a while, but this is a hot off the press paper just came out looking at chronic pancreatitis and how shear wave elastography is performing even better than standard elastography and contrast-enhanced EOS. So this may be coming out more in the next couple, three years. And I just wanted to highlight this in today's session before we close out this talk. So do advanced EOS platforms offer value? I think that's what you want to know from this talk. Well, the value lies in the eyes of the beholder. I think this is certainly a quest for better resolution, higher accuracy, solving conflicting dilemmas in clinical diagnosis. Will it replace pathology? I'm not sure yet until other paradigms change and the technique gets a little bit more reliable, more widely adopted, there are better studies and so forth. But it certainly does help in certain scenarios, just like molecular markers help with cyst analysis and so forth. Certainly those patients who are unable to get high quality cross-sectional imaging and have had repeated FNA attempts still remain a concern. I think it's valuable to use it. So more to come on that. Final polls, elastography and CEUS are definitely complementary. They're not yet widely available, but are increasingly becoming common, especially in the new systems. The diagnostic accuracy of FNA and FNB are still very high and consistent, especially in expert hands. And so I think it may have a role in indeterminate lesions, but mostly may have a role in differentiating autoimmune type of benign conditions from pancreatic cancer.

case-based presentations

mediastinal EUS

fiducial placement

elastography

contrast-enhanced EUS

diagnostic evaluations