So, the topic of my presentation is the management of pancreatic fluid collections. I have no conflicts of interest to disclose. So, pancreatic fluid collections are common, occurring in up to 40% of cases of acute pancreatitis, though the good news is that the large majority are of no clinical consequence and will resolve spontaneously. Acute collections that are more than about 5 or 6 centimeters in size are those associated with severe and necrotizing pancreatitis are the ones that tend to persist and may ultimately require drainage. And as mentioned by Anne-Marie, it's always important in this process to exclude a pancreatic cystic neoplasm because percutaneous or endoscopic drainage of a neoplastic cyst can result in significant iatrogenic or procedural and oncological complications. The last thing you want to do is rupture one of those cysts and have those neoplastic cells spill all over the peritoneum. And as mentioned, this is particularly important in a patient with no prior history of acute or chronic pancreatitis, but it is important to keep in mind that pancreatic cysts are sometimes the cause rather than the consequence of acute pancreatitis if they are neoplastic. So, just because a patient has had acute pancreatitis doesn't automatically mean that the cyst is inflammatory. It could be neoplastic and having caused the attack. So, along the treatment algorithm for pancreatic food collections, the first question or the first order of business is to establish to the greatest extent possible that the cyst is actually inflammatory. And the second step is to determine whether drainage or treatment is actually indicated. And the guiding principle here is that the risk of an iatrogenic complication related to any form of drainage in the large majority of cases will far outweigh the risk of a spontaneous complication like intracavitary bleeding, infection, or spontaneous rupture of the cyst, which occurs, but is very uncommon. So, these old school principles that many of us learn that if a cyst is more than six centimeters in size, if it's been around for more than six weeks, if it's growing rapidly, these are all classic indications for drainage, that whole paradigm is now outdated and no longer applicable in clinical practice. Instead, regardless of size or anything else, cysts should be drained, fluid collection should be drained if they're causing significant symptoms, if they are infected, or if they're causing compression of a critical structure like the bile duct, an important blood vessel, or the GI tract, in which case they would likely be symptomatic anyway. And when thinking of and considering the best drainage strategy, I do think it's worthwhile reviewing the revised Atlanta classification nomenclature for pancreatic cysts, which divides cysts according to maturity and content. So, cysts that are present within four weeks of an attack of pancreatitis are considered immature for the most part if they're an acute, and they're an acute peripancreatic fluid collection if they contain pure fluid, or an acute necrotic collection if they contain pancreatic or peripancreatic necrosis or necrotic debris. On the other hand, collections that are more than four weeks out from attack of pancreatitis are considered mature, and this is the nomenclature with which we are more familiar, so pancreatic pseudocyst if it's pure fluid, and a walled-off necrosis, or what used to be called an organized necrosis if there's solid necrotic content in the body. And so, once we determine that treatment is indicated, that the patient is symptomatic, or there's an infected cyst, et cetera, then the next fundamental question is have we reached that four-week mark? Because before four weeks, the wall, again, is considered immature in the large majority of cases, and is less likely to withstand puncture and dilation or suturing in the case of surgical intervention, and importantly, the process is in evolution. So, even a surgical necrosectomy may not be effective because the patient could develop additional necrosis in the subsequent days. And so, these are a couple of examples of pancreatic fluid collections in evolution in the early phase. As you can see, early on, there is some change and expansion, but no clear evidence of a mature capsule around the cyst cavity. It's not till week five that you see this sort of homogeneous and thick rind around the entire lesion indicating that it's mature and potentially ready for endoscopic or surgical drainage. So, for these reasons, and also because any intervention in the acute phase can worsen the systemic inflammatory response, the goal is to temporize treatment to the greatest extent possible. And this can be done with IV pain and nausea control, it can be done with deep enteral tube feeds if the patient is intolerant of PO intake, and with intravenous antibiotics if there is suggestion of infection of the pancreatic cyst. And the number one reason for failure of conservative management in this early phase is refractory infection, which should ideally be addressed via percutaneous drainage according to what's known as the step-up philosophy. So, now we know on the basis of robust observational and randomized control trial data that a strategy of stepping up initially to percutaneous drainage if conservative management has failed, and then if percutaneous drainage fails, stepping up to minimally invasive surgery reduces mortality and improves overall outcomes related to the traditional strategy of early surgical necrosectomy. And so that's known as a step-up approach. And in the early phase, within four weeks, if conservative management has failed, percutaneous drainage is indicated whether or not solid necrosis is present within the cyst cavity. So it's indicated for acute peripancreatic fluid collection as well as an acute necrotic collection. However, if there's reason to believe that there's solid necrosis within the cavity, then drainage would ideally be achieved through a retroperitoneal tract. And that's because, and this is certainly for collections that replace the body or in the region of the body and the tail of the pancreas. And that's because if the patient fails to improve with percutaneous drainage, and the goal is to step up to minimally invasive surgery, then the best operation in this context is what's known as a VARD, or a video-assisted retroperitoneal debridement. And a VARD makes use of this retroperitoneal drain tract to allow mechanical debridement of the necrotic tissue using laparoscopic instruments. For collections that are right of midline, so those that are associated with the head of the pancreas and sometimes they end up in the right upper quadrant, then the best minimally invasive operation in this context is a transgastric, ideally laparoscopic necrosectomy. And the conventional open necrosectomy that opens the abdomen and exposes the lesser sac, there's really no role for that in clinical practice unless the minimally invasive options have been exhausted and the patient is basically in dire straits. So beyond four weeks, then we're dealing with a mature collection, either a pseudocyst or a walled-off necrosis, and at this stage in the game, endoscopic options become viable. And that's because at this stage, the cyst wall has likely become adherent to the GI tract, and importantly, the capsule has become mature enough to withstand puncture and dilation and manipulation and so forth. But after four weeks, the distinction of whether it's pure fluid or has necrotic content, so the distinction between pseudocyst and walled-off necrosis becomes much more important because it has concrete implications on the drainage strategy. The problem is that CT scan, unfortunately, is inadequate to differentiate pseudocyst from walled-off necrosis. There's nothing specific about the appearance of the fluid that will suggest definitively that there's solid content within. Now there are some secondary findings on CT scan that can help differentiate the two, and in my experience, the most helpful is when you see a collection that replaces a portion of the pancreas and leads to discontinuity of the gland. And what this implies is that this portion of the pancreas has died, and that dead necrotic tissue has been contained within this new cavity. And so this is an example of that phenomenon, and you can see this collection replacing the sort of downstream body of the pancreas, and on the basis of the scan alone, you really can't, there's nothing about the Hounsfeld units or anything else that tells you with a high level of certainty that there's necrotic content. But on EUS, it's very obvious that it's mostly fluid, but there is this hyper-echoic shadowing material, and so there is some solid content. So in this context, if it's gonna make a decision in clinical decision-making, the distinction should be based on EUS or MRI, both of which are much more effective than CT scan. So for collections that contain pure fluid in this late phase, then the optimal drainage strategy is based on the size of the collection. And collections that are less than five centimeters in size are best drained via primary trans-papillary drainage, in other words, ERCP with a pancreatic stent. And the reason that is is that small collections tend to A, respond really well to this, and secondly, often they don't provide sufficient working space for safe and effective transmural drainage. And the goal of trans-papillary drainage is to re-divert the flow of pancreatic juice down a new path of least resistance into the duodenum rather than out the defect into the collection, which will ultimately allow the collection to collapse and heal. And the goal, of course, is to bridge the papilla, which is a high-pressure zone, and is often sort of where the pancreatic juice gets hung up and gets re-diverted. But in many cases, there also is a stricture downstream of the collection, and it's important to recognize that, particularly with chronic pancreatitis and to drain that as well. Now, as opposed to bi-leaks, in which eliminating trans-papillary pressure with a short stent or a biliary sphincterotomy is more than enough to drain the leak, sorry, to treat the leak, there's no need to seal the leak per se. In the pancreatic duct, there are observational data that suggest that if you bridge the leak, that might be associated with improved clinical outcomes, and perhaps this is because pancreatic juice is just much more caustic than bile. For collections that are larger than five centimeters, then the optimal treatment approach is transmural drainage, what's known as a cys-gastrostomy or cys-duodenostomy, plus or minus ERCP. And the goal of transmural drainage is to create a connection or an asthmosis between the cyst cavity and the GI tract, and this allows cyst contents, the fluid, to drain out into the lumen of the GI tract, and subsequently the cyst can collapse and heal. And this, of course, had been done traditionally, and even nowadays, although it's almost always done laparoscopically, there's still about a 20% risk of serious adverse events and a mandatory hospitalization. And endoscopic drainage had been introduced in the mid-90s, but in the last 15 years or so, has gained a lot of traction, particularly with the advent and diffusion of EUS, which makes the procedure much less risky. Most people would agree that in 2021, and even for the last decade or so, endoscopic drainage of pseudocysts that are anatomically amenable has supplanted surgical drainage. The technical sequence is that an FNA needle under EUS guidance to exclude interceding vessels is advanced into the cyst cavity. Through the FNA needle, a wire is advanced into the cyst and coiled, and then over the wire, the tract is established and expanded, typically with a dilating balloon, and then one or more stents are placed across the tract to promote drainage of cyst contents into the lumen of the GI tract. And this is an EUS image of a completely anechoic, so a completely black cavity. So this is a pseudocyst, no necrotic debris. A needle and a wire under EUS and Doppler guidance have been advanced into the cyst cavity, and under fluoroscopy, you can see that the wire has been coiled in the cavity. And then over the wire, the tract is established, in this case, initially with an ERSP catheter, and you can see copious flow of pseudocyst fluid into the lumen of the duodenum. And then over that wire, the tract can then be expanded. In this case, we are inflating a biliary dilating balloon in order to fully establish that tract. And you can see, once the balloon is deflated, an additional purulent fluid will drain into the duodenum. And then, as mentioned, one or more stents can be placed across the tract to promote ongoing flow and eventual resolution. Guess I need a little... So traditionally, the tract was maintained using plastic double pigtail stents. So this is an endoscopic view. You can see a single pigtail here in the lumen of the stomach. And under fluoroscopic guidance, you can see a second pigtail within the cyst cavity. More recently, many people have moved toward using fully covered self-expanding metallic biliary stents for this indication. You can place the stent across the tract as you would the plastic stent. The advantage of the fully covered metallic stent is that it's more efficient and more straightforward to deploy. And in principle, it's less likely to occlude because of a larger lumen. And so there are some early observational data that suggests that certainly the efficiency of the procedure's improved by placing a metallic stent and perhaps slightly better clinical and technical outcomes. And the most recent and perhaps most exciting innovation in this space, although it applies, in my practice at least, much more to walled-off necrosis than it does to pseudocyst, is what's known as a lumen-opposing metal stent. And so these are fully covered self-expanding stents that have these large anchoring flanges on either side of the saddle, and they can have a large lumen up to two centimeters in size. And the version that's available in the United States is delivered on an electrocautery-enhanced platform that allows near-simultaneous puncture of the cyst cavity, establishment of the tract, and deployment of the stent. So it can make the procedure very efficient. So if you look at the literature in aggregate, endoscopic pseudocyst drainage is associated with a greater than 90% technical success rate, a slightly lower clinical success rate, which implies long-term resolution of the cyst cavity. And this has to be balanced against about a 10 to 20% risk of adverse events, although the good news is that almost all of these adverse events are addressable endoscopically. Typically they pertain to secondary infection of the cyst cavity. And on the basis of observational data and a randomized control trial, endoscopic and surgical drainage appear to be clinically equivalent, but the endoscopic approach is associated with lower cost and a shorter hospital length of stay. And many of these cases are done in the outpatient setting anyway, and so there really is no length of stay. And as mentioned, endoscopic drainage in this context for this type of cavity type have supplanted surgery. Now the more difficult clinical problem is the Waldorf necrosis, and that's because in addition to pure fluid, these cavities contain thick, solid necrotic debris, which prevent the cavities from collapsing and also increase the likelihood of occluding transmural stents. So for that reason, many of these, but not all will require what's known as a direct endoscopic necrosectomy, and that's the process of advancing the scope through the tract, in this case, in this illustration, through aluminum opposing metal stent, into the cyst cavity, and then we'll use a multitude of instruments to physically grab and debris that necrotic debris and deposit it in the stomach. And this is a fluoroscopic view of a gastroscope being advanced through, this is aluminum opposing metal stent into the cyst cavity. And this is a brief video example of that. This is a necrosectomy through aluminum opposing metal stent, which by the way is not mandatory for this procedure. We did endoscopic necrosectomy for many years through a fresh tract, but it probably does add an element of efficiency and safety. As you can see, this is occluded by necrotic debris, and we're gonna use a coin grasper to sort of move that out of the way, which allows all this purulent fluid to sort of decant into the lumen of the stomach. And then we're gonna enter the cavity where we'll encounter a large amount of necrosis, and as mentioned, we'll use several different kinds of endoscopic instruments to remove all that necrosis from the cavity, which will eventually allow it to collapse and to heal. In this case, we're using what's known as a three-pronged grasper to grab that necrosis and bring it into the stomach. But in my practice, by far the most useful instrument is actually a standard hexagonal polypectomy snare. It's very safe and forgiving against the cyst wall, and you can grab very large pieces of necrosis. And if the piece is too large to actually sort of milk it through the tract, then you can deliver electrocautery to transect the necrosis, as we've done here, and then we'll use that same snare to grab the morsel and bring it into the stomach. And this procedure is definitely a labor of love. These are all, of course, repurposed instruments, so it can take a long time. It's definitely effort and resource intensive, and sometimes we do several sessions over the course of many weeks to months to be able to get all the necrosis out. So the worldwide experience with endoscopic necrosectomy has increased substantially in the last decade. And the literature varies, but the technical success rate is reported to be somewhere between 70 and 95%, with a slightly lower clinical success rate, although the complications of endoscopic necrosectomy are higher, clearly higher than pseudocyst drainage, including the small but real risk of AR and CO2 embolism. As mentioned, these are technically complex procedures, and they're quite resource and effort intensive. And so, whereas for pseudocyst, endoscopic drainage is clearly the dominant strategy, it's still a rather big question for walled-off necrosis. Having said that, there are now three randomized control trials comparing endoscopic to surgical necrosectomy, comprising 190 patients, and in aggregate, these studies appear to suggest, and they're all, of course, exploratory and preliminary because they're small, but they do suggest improved outcomes including less pancreatic or duodenal fissurely, better cardiovascular outcomes, shorter length of stay associated with endoscopy, suggesting that perhaps it should be first-line therapy, not reserved for suboptimal surgical candidates, as it had been, certainly, over the first phase of my career. Now, of course, there is a major role for surgery when it's a very complex collection, particularly enormous collections that track down both paracolic gutters. You can consider surgery, certainly in a young, surgically fit patient, especially if they have a disconnected pancreatic tail, in that context, it may not make a lot of sense to do multiple necrosectomies over the course of weeks and then have them end up in the operating room anyway for a distal pancreatectomy. And, of course, we all see patients in whom the progress, the improvement, is not rapid enough, so they're failing to thrive despite multiple necrosectomies. That's a situation that was mentioned earlier. You gotta swallow your pride and involve our surgical colleagues. So, again, as opposed to pseudocysts, this very much remains a multidisciplinary disease, but, admittedly, we don't discuss every one of these with our surgical colleagues. As mentioned, complications are more common with walled-off necrosis than they are with pseudocysts, and by far, the most common complication for both is cyst cavity infection. This is usually related to stent dysfunction, so migration or occlusion of the stent with worsening infection in the cavity, but can occasionally, with walled-off necrosis, be because of a jailed-off compartment that is just sort of full of concentrated pus, and these are almost always addressable with IV antibiotics, repeat endoscopy to replace the stent, do additional debridement, et cetera. Bleeding is a significant risk. This can be because of a pseudoaneurysm, or it can be venous. The dwell time of the lumen-opposing metal stent has recently been implicated in bleeding, and so most of us will try to get the lumen-opposing metal stent out within about three or four weeks. Pseudoaneurysm bleeding is dramatic, and it's scary, but it has a great solution, which is angiographic embolization. Venous bleeding, on the other hand, can be a huge nightmare, particularly if it's a major vessel like the portal vein, for example, if the cavity gets into the portal vein, and these do not have great solutions, so this is actually the more dreaded outcome. Perforation can occur, but if it's contained in the lesser sac, if the patient's doing well along some of the principles that Olaya mentioned earlier, surgery's not always required, and of course, if an ERSP is done, then pancreatitis is always a risk. For the sake of completeness, but also recognizing that I've run over, I just want to briefly mention that there are several additional questions and considerations around transmural drainage and endoscopic necrosectomy, including whether every single patient, or even most patients who have transmural drainage also need an ERSP with stent placement. This is an area of controversy with strong arguments in both directions. There is some controversy around whether every patient who undergoes endoscopic necrosectomy, who undergoes transmural drainage, needs an upfront necrosectomy, which is my practice, versus stepping up to direct necrosectomy if stents alone don't do the trick. There's an emerging role for adjunctive strategies, including hydrogen peroxide, and of course, we still don't know the best treatment algorithm for patients with a disconnected pancreatic duct, whether that's destination transmural stenting to chaperone the flow of pancreatic juice from the tail into the stomach in the long run, for which there are good long-term data from Europe, versus an automatic distal pancreatectomy. So with that, I thank you very much for your attention.

pancreatic fluid collections

drainage

conservative management

endoscopic drainage

surgical drainage

complications