Our next speaker is going to be Professor Michelle Anderson. Dr. Anderson is Associate Professor of Medicine in Gastroenterology and Clinical Chief of Endoscopy at Michigan Medicine. Her clinical practice is predominantly on pancreatic oblivion endoscopy and excellence in endoscopic ultrasound. She's been associated with the ASG for long and has been the ASG Distinguished Service Awardee recently. Dr. Anderson, we're looking forward to your talk on EUS for chronic pancreatitis. Thank you. Thank you, Nagi, and I would like to thank the course directors and ASG for inviting me to participate in today's events. Certainly we could spend an hour talking about the role of EUS in chronic pancreatitis. I've been given 13 minutes, so we're going to have a more focused discussion. I have a couple of disclosures, as you can see on the screen in front of you. So the topics that the course directors did ask me to focus on today include EUS versus MRI for the diagnosis of chronic pancreatitis and EUS for celiac plexus block and chronic pancreatitis. So starting with the first topic, EUS versus MRI, which is better for the diagnosis of chronic pancreatitis? So I would argue that either of the tests, EUS or MRI, are adequate when we're talking about either extent of the spectrum in terms of disease, whether that's a normal pancreas, as you can see here in the MRI and EUS, or in the case of advanced disease, again, MRI versus EUS. The real difference and where there may be improved performance for endoscopic ultrasound is when we're talking about early chronic pancreatitis or minimal change chronic pancreatitis. This is an MRI from a patient that has some atrophy as well as some main pancreatic duct irregularity, and this patient was read out on the MRI report as possibly having early chronic pancreatitis. The EUS showed similar changes with atrophy and hyperechoic strands. So for those that don't do endoscopic ultrasound, it's important to note that we do use a set of criteria. We just don't put the scope in and look around and decide if we think subjectively if the patient has chronic pancreatitis, and the criteria we use are the Rosemont criteria, which were developed by a group of international experts in endoscopic ultrasound who met at the O'Hare Airport, which, by the way, is in Rosemont, Illinois, which is where the name comes from, and that was back in 2007. And so they agreed on different criteria that could be evaluated during endoscopic ultrasound, and they divided these into parenchymal changes and ductal abnormalities. And then there's a grade of the confidence that we have when these different features are seen. So calcium trumps everything else, and that's a major A criteria, whether that be in the parenchyma with hyperechoic foci and shadowing or in the duct in the form of main pancreatic duct calculi. And then we have minor criteria, and honeycombing is a major B criteria. And so we add up all these things at the end of the procedure, and we decide where do they fit in this right-hand table in terms of our confidence in diagnosing chronic pancreatitis. This is a really useful slide. We keep up a copy of this outside of both of our interventional rooms, because even after years of using this, I still haven't memorized all of these, and I use this table almost daily to decide where the patient falls in this spectrum. So I think it's really important before we go on to a discussion about abnormal pancreases for everyone to see what a normal pancreas looks like. So this is an EUS exam in a patient that was having an EUS for a non-pancreatic indications, and what you can see here is just a beautiful, homogeneous, smooth pancreas. No hyperechoic foci, no dilated duct. The duct measure is 1.3 millimeters, and so I want you to keep this in mind as we go forward and I show you some of the cases that I'm going to present, because there really is a difference. So our first case is a 48-year-old female with three years of epigastric abdominal pain and an elevated lipase that was about one and a half times the upper limit of normal. And so for reference in this upper left-hand corner, you can see that normal pancreatic EUS that I just showed you. This is the EUS from the patient I just presented, and so you can see this is definitely not like this one. We've got hyperechoic strands, we've got some dilated side branches, and so this patient had two minor criteria, and so this would fall into the Rosemont category of normal. And so this patient, I would say, it's not perfectly normal, but you don't meet criteria, and I would really avoid labeling this patient as chronic pancreatitis, because I think we do a disservice to patients when we label them with these minimal criteria, and you could have changes like this just from smoking or excessive alcohol consumption. So our next case is a 56-year-old male with symptoms of malabsorption, including bloating, weight loss, and low-level abdominal pain. The patient actually had a CT scan from his primary gastroenterologist, which showed a cyst in the proximal body, and this is the EUS from that patient. You can see hyperechoic duct walls right here, lots of hyperechoic strands, dilated side branches, and even some lobularity. And so the next thing we're going to show here is the cyst. We did do an FNA of that, and the cyst fluid was high in amylase, greater than 14,000, and had unmeasurable CEA consistent with a pseudocyst. And so this patient, unlike the last one, I would talk to and say, you know, it really does look like you have chronic pancreatitis. These malabsorptive symptoms are probably PEI, and the patient should be tested for that and treated accordingly. Our third case is a 64-year-old male who had abdominal pain, and his primary gastroenterologist did an upper endoscopy because of that pain and saw a subepithelial gastric mass. And so this is the EUS from that, and we start off with what is a very classic gastrointestinal stromal tumor arising off of muscularis propria here. Because of the pain, we looked at his pancreas, and lo and behold, this patient has a main pancreatic duct calculi in the center of a dilated duct. He had irregular pancreatic duct contour, dilated side branch, hyperechoic strands, and this was one major A and four minor criteria, so completely consistent with chronic pancreatitis. There is no question that this patient has chronic pancreatitis. So many have criticized the Rosemont criteria for over-calling chronic pancreatitis. It turns out from studies, one from our institution and another institution, have shown that the Rosemont criteria are actually pretty conservative compared to the conventional criteria that we used to use and may, in fact, underestimate chronic pancreatitis prevalence. If you're seeing patients that have isolated elevations in amylase and lipase and you're not sure, those patients probably should be evaluated with endoscopic ultrasound because studies have shown that they may very well have underlying chronic pancreatitis. A study from Butani and colleagues also showed that the number of Rosemont criteria actually correlates with subsequent need for surgery, suggesting that these are valid and legitimate criteria that are being used and do have clinical impact. Finally, I would point your attention to a recent ASGE practice guideline on enhanced imaging in EUS, and this is a really hot and interesting field. Many believe that enhanced imaging will make EUS more sensitive and more specific in the arena of chronic pancreatitis diagnosis. So back to our original question, which is better? Unfortunately, the data comparing EUS to MRI are sparse. The first study I'm showing you here is actually from Mayo-Jacksonville. This involved roughly 100 patients where they compared EUS to MRI for the diagnosis of chronic pancreatitis, and they used a composite gold standard of ERCP surgery or long-term follow-up. In this study, roughly the median follow-up was 16 months, and what you can see here is that EUS was significantly more sensitive than MRI, but about the same in terms of specificity. In my opinion, this is the best study out there that did a direct comparison. A subsequent and more recent systematic review and meta-analysis, which included this study and 42 others, showed that the sensitivity and specificity of MRI and EUS are pretty equivalent and without statistical difference. So what do I do? Really it depends on how that patient is presenting and why I'm considering a diagnosis of chronic pancreatitis. If the patient presents with malabsorptive symptoms alone, I'm going to start with stool studies. The caveat here would be if cancer is in the differential, then I'm going to also do imaging of some form, whether that's EUS or CT or what have you, or if there's a new unexplained pancreatitis in this patient. If their stool study suggests pancreatic enzyme insufficiency, you'll treat accordingly. If it doesn't, you'll look for alternative causes of malabsorption, and if there are none, then you'll consider alternative diagnoses. In the patient that presents with pain, with or without malabsorption, your first-line test should probably be CT, and that's because it will pick up the vast majority of severe chronic pancreatitis and normal pancreas. Caveat here again would be that you should use EUS first if there's cancer in your differential diagnosis or new unexplained pancreatitis, especially in patients over the age of 40. If you're still suspicious and you haven't come to the diagnosis, I think MRI and EUS are pretty equivalent, and you would choose one over the other based on local expertise and patient preference. I'm going to finish by talking briefly about EUS for celiac plexus block in chronic pancreatitis, including indications, outcomes, and how to do it. As a reminder, the area that we're targeting when we do these blocks is actually the celiac plexus that surrounds the celiac artery, the first major artery off the aorta below the diaphragm. This is done with a linear scope transgastrically by injecting in that general area. Indications for this include unresectable pancreatic cancer, chronic pancreatitis, pelvic cancer with intractable pain, and then there's a smattering of other case reports for other indications in the literature. Contraindications include an inability to tolerate EUS, coagulopathy, and then there's a relative contraindication of tumor encasement of the celiac trunk, and that's because when that occurs, it makes it very difficult for us to identify our targets and to safely inject the pancreas, the celiac ganglia, excuse me. If you do this, you probably have a protocol. If you're thinking about doing this, I would strongly encourage you to have a protocol for your unit. This is a picture of ours, and we all use it, the staff, the techs, the attendings to get our equipment together and make sure we're doing it the same way every time. I'm happy to share that with anybody that might want to look at ours. We have our patients arrive an hour before their appointment to receive a liter of normal saline to counteract the very common orthostatic hypotension that can occur afterward. The patients get their informed consents, and we sometimes use antibiotics. The cases we use antibiotics in would be those that are getting blocks. Neurolysis, we don't use antibiotics because alcohol is bacteriostatic, and there really probably isn't much benefit in that setting. So here's our equipment. If we're doing neurolysis, of course, we're going to use dehydrated alcohol and bupivacaine. If we're doing a block, then we're using the bupivacaine plus a steroid, in this case, Canalog, various syringes, and our injection needle. So there are needles that are specifically designed for celiac plexus neurolysis. We generally don't use one of those. At Michigan, we use a 19-gauge or a 22-gauge needle, as I showed you in the last picture. So what does this look like? This is a video from a patient that I did a celiac plexus neurolysis in several years ago, and we can see the aorta and longitudinal section, the celiac artery. We identified a ganglion on the right side. We puncture the ganglia. We check for blood by aspirating. We inject first with the bupivacaine and then with the dehydrated alcohol, and you'll see this bright, sparkling halo that comes out. That always happens with the dehydrated alcohol. It doesn't happen with the steroids, and that's an effect of the alcohol in the tissue. We turn the scope and identify a left ganglia, do the same thing on the left side, and we always inject saline on the way out, and that's so that we don't trail dehydrated alcohol through the gastric wall or the diaphragm. Very important. So there are various techniques for doing this. You can inject bilaterally. You can inject unilaterally with all the medications on one side, usually the left. You can inject directly into the celiac ganglia, as I just showed you, or you can do broad neurolysis where you inject at the level of the celiac artery and the level of the superior mesenteric artery. There are no studies which compare all of these different techniques. There are some that compare one against another. My personal opinion is that if you can see the ganglia, which I see less than 50% of the time, those are the best to target. And secondly, I think the second best one is a broad neurolysis injecting both celiac and SMA, but it's often difficult because they're in the same plane, so it's really hard sometimes to get anywhere close to the SMA without going through the celiac bilateral. And then even if you have to do unilateral, it's still fine, and it will likely work. So what about outcomes? Unfortunately, the data for celiac plexus block in the setting of chronic pancreatitis is not nearly as good as it is for celiac plexus neurolysis in the setting of cancer. There are zero placebo-controlled trials. There are two systematic reviews and meta-analyses. One of them I'm showing you here. And what you can see here from the forest plot is that roughly 50% of patients had some benefit from a celiac plexus block in the setting of chronic pancreatitis. In this review, they had a median duration of only 11.3 days with a range of zero to 37 days, so really disappointing, not very strong evidence to do this, and it's not something that I offer to all patients. I will say that there is data showing that if a patient responds to a celiac plexus block they're likely to respond to one again later. If they do not respond the first time, they are not going to respond the second time, so don't expect that that's going to happen if you bring them back. Quickly, complications are common and you need to share this with your patients. These ones on the left occur in 20 to 30% of patients, including orthostatic hypotension, acute increase in pain or diarrhea or incontinence, which is usually temporary. There are also some rare and sometimes catastrophic complications that can happen as well, and those also need to be at least mentioned to your patient when you do the consent. So my take-home points are use CT first to diagnose chronic pancreatitis and reserve EUS or MRI if equivocal and high suspicion for chronic pancreatitis. Avoid labeling a patient with minimal criteria as having chronic pancreatitis. There is weak evidence for benefit of celiac plexus block in chronic pancreatitis, and if offered, you should advise patients on the risks and benefits. Thank you. The first question is, if you have a patient with suspected chronic pancreatitis and abdominal pain, would you do an EUS first or would you send them for an MRCP and CT and then call them back for an EUS? Yeah. I mean, I think if they have a history of pancreatitis or they have risk factors for chronic pancreatitis, like smoking and alcohol use, that might be a person that I might consider doing EUS first in, but if those are not present, then my go-to is going to be CT scan. Okay. Well, it's a fantastic talk as usual. I have a question about, it seems like I see all the patients with equivocal CT scans and EUS that's also suggestive with negative stool samples. Two questions. One, what type of stool samples are you actually looking at specifically? And two, what do you do with that patient? Yeah. I mean, that's like one of the examples that I presented here, kind of, you're on the fence. And so the stool studies that I send is I send a stool fecal elastase and I send a spot fecal fat. And the reason I do that instead of a quantitative study for stools as a first line is I can get all of that information with one sample and it's less onerous and the patient is likely to complete it. If the fecal fat is positive and the fecal elastase is low, then I'm confident that that patient has chronic pancreatitis. If the spot fecal fat is not positive, then that's the patient I might do a 48 or 72 hour collection in. In the setting that you just described where there's a equivocal CT or even a normal CT, but good history and some abnormality on EUS, those patients I hedge. I say, you know, I think you might have this, time will tell. And I use that as an opportunity to get people to have better lifestyles. If this is what you're looking at now, you've got to stop smoking and you've got to moderate your alcohol use and we'll see what happens. And Michelle, one final question is, how long would you wait to do an EUS after acute pancreatitis? That's a really excellent question. So yeah, so in general, we try to wait four weeks. The only time that we might push that is if we have a really high suspicion for pancreatic cancer. And the reason for that is obvious, but generally you should wait four weeks or more because you can be fooled. The changes that are associated with the acute pancreatitis can look like chronic or even like cancer when it's not really cancer.

endoscopic ultrasound

diagnosis

chronic pancreatitis

MRI

accuracy