This meeting is being recorded. Hi, my name is Luke Kaplan. I'm a surgeon at Penn and SCCM's immediate past president. Today, we'll talk about acute pancreatitis. Our objectives will be to go over presentations and etiologies, look at staging and diagnostic tools, talk about early enteral nutritional support, discuss antibiotic courses, and determine the timing and nature of what kinds of interventions are most appropriate for your patient. Acute pancreatitis has a country-dependent incidence. 5 to 80 per 100,000 is a very broad range. Nonetheless, alcohol, gallstones, injury, and hyperlipidemia predominate as the four major causes. Presentation severity is not uniform, but up to a quarter have a life-threatening course. There are two phases that are described with regard to pancreatitis-related mortality, early and late, and you can see the descriptors there. Mortality trends, if you think about almost a million patients, this is crude mortality per year. You can see that during the time that this was evaluated from 2008 to 2015, mortality is in fact better. This is in large part related to improvements in critical care, but also in therapeutic interventions. But to get there, you have to render the diagnosis. The diagnosis has very typical history and physical examination, but you need two of three criteria. Symptoms consistent with pancreatitis, GI dysfunction, back pain, a serum amylase or lipase that is more than three times normal, and some imaging that supports the putative diagnosis, which the following would suggest severe acute pancreatitis. I won't give you the answer, but these are things for you to work through and pay attention to as we talk about it. Serum amylase of 5200, Apache 2 of 7, and a C-reactive protein of 100, CT scan, image pancreatic necrosis of 25% of the total body, tympanic membrane temperature of 40 degrees Centigrade, and an ABG PaO2 of 58 on 50% FiO2. Think about how each of these would interface with the diagnosis of pancreatitis. There's a classification that can help you. This is the revised Atlanta classification. Our current president is in fact from Atlanta. Mild acute pancreatitis, no organ failure, and no local or systemic complications, moderately severe pancreatitis, less than 48 hours worth of organ failure, and or some local or systemic complications, but nothing that is persistent, and then things that persist where you have single or multi-organ failure, we call that severe acute pancreatitis. Etiologies we've spoken briefly about, biliary pathology and alcohol top this particular list, but it can include anything from infection to snake venom. There are occasionally pancreatitis etiologies related to pancreatic stone protein and other causes that we have not yet discovered, but those are far less common. There is a cascade that's been well described. There's acute inflammation with a capillary leak. There is hypovolemia. It leads to shock and acute kidney injury and has significant overlap with acute lung injury and ARDS, all from acute pancreatitis. There are a number of relationships, whether you have an endogenous or an exogenous trigger of pancreatic inflammation. Interleukins, TNF, and nucleofactic kappa-beta have interface with cells and the endothelium. There is an endotheliopathy that goes along with this that traffic neutrophils to the site of inflammation and exaggerates cell-mediated injury. Time course is presented here. You can see ED presentation in orange, which spans a day and a half after the inflammation begins. Cytokine production is fairly persistent and looking at up to 96 hours remains above baseline. And it's about that 60-hour mark that most organ failures begin to present themselves. C-reactive protein, in fact, has predictive value. What you're looking at is, in the upper left-hand corner, CRP levels, mild, moderate, and severe. They're color-coded, and you see they're nicely broken out here, mild, moderate, and severe, over 24 to 48 hours. If you look at the interval change in CRP that also describes the severity, you can see that those with most severe infection, here in gold, have more change at any of those measured time intervals. So that you can compare the Atlanta classification severity that we just went over with mean changes in CRP. You can see at 24, 48, and 72 hours, the largest change correlates with the highest Atlanta severity. And so there are 48 hours. A delta of 90 is equivalent in terms of predictive value of an absolute concentration of 150. And a new cutoff for the moderately severe to severe has been listed as 190. Question two, 56-year-old, which is now unfortunately younger than me, presents with new onset pancreatitis. Some of the labs are, what you see here, hypoxemia, leukocytosis, hyperglycemia, and a relatively low calcium. How do you manage this person? And the big divider is hospital admission versus ICU admission, and whether you support them in terms of resuscitation first, or whether they should undergo some kind of diagnostic evaluation. Think about what you do, and think about what the most appropriate is, and whether you have a pathway or a protocol at your institution. There are the age-old Ransom criteria, assuming that you have no gallstones. And they're listed here at admission and during 48 hours. That correlates very nicely with that change in CRP. You can see all of these descriptors, and the more Ransom's criteria you have, the higher is your mortality. Hypocalcemia is incredibly common, often related to saponification of fat within your intra-abdominal space. And so when do you image? With mild pancreatitis, you don't really need any. Perhaps you get an ultrasound as a baseline, or look for other kinds of conditions, but it's not really required. Contrast-enhanced CT is indicated for moderate or severe acute pancreatitis, if you're not sure that's what you're dealing with, or for people that don't improve after 72 hours because you're looking for complications, or those who fall off the pathway and become unstable. For most patients, you don't need to scan them before 72 hours because you're looking for complications rather than establishing the diagnosis. There are some terms that define complications that are worthwhile to understand, and it spans everything from edema without necrosis as interstitial edematous pancreatitis, to pancreatic pseudocyst, there's a time frame in here, more than four weeks, to acute necrosis, fluid and necrotic material within that particular portion of necrotizing pancreatitis, within the first four weeks, but no well-defined wall, and walled-off necrosis. Each of these has a mortality price to pay, and it should also define a therapeutic pathway. And so when you're looking at things, this is a well-described and well-defined pancreatic pseudocyst, homogeneous fluid-like content, and a well-described wall, not much inflammation around it. Pancreatic necrosis, now looking at what used to be the pancreas, there's now replaced with this fluid-filled collection that is in fact in the pancreatic bed, this is necrotic pancreas. And you can define the extent of necrosis when you're looking at all of that, because this is the necrotic portion, and there's the tail, which is in fact still alive. When you think about the impact of necrotizing pancreatitis on organ failure, you can correlate the pancreatic necrosis extent, less than 30, 30 to 50, and more than 50, overall is in blue, sterile necrosis, or infected necrosis, and you can see at every proportion of pancreas that's involved, infection increases the likelihood of organ failure. This is a well-known series of events. So what are their controversies? Because diagnosing it is not particularly controversial, and the fact that more necrosis and infection is also not controversial with regard to complications and mortality. Feeding them happens to be. When I was training, we never fed patients with pancreatitis other than using TPN, and no lipids were included because hyperlipemia also can trigger pancreatitis. So we think about that Atlanta classification mild that Brackley resolves. They don't need any support. They can nourish themselves, severe or prolonged. They have a four to six week period, and it's in this group that you have to decide enteral versus parenteral support. If you choose a parenteral route, you will most certainly experience glycocalyx failure as the villus tips atrophy, and that entire brush border enzyme layer that is required for complex nutrient processing and absorption will also fail. This is a 72-hour process. So you have to decide, do you need to support it, and can you support it in your patient? I think the former is yes, the latter will depend upon your unique patient. But if you can feed them, enteral outperforms parenteral with regard to every single metric that you can identify, and here they're presented with regard to mortality, infection, the need for intervention, and organ failure. Creating that brush border enzyme system takes effort and energy, and therefore maintaining it is only supported by luminal nutrients, not IV ones. Question three, 42-year-old woman with severe acute pancreatitis of unknown etiology. Which of the following are the most important initial investigations? Imaging the biliary tract, right upper quadrant ultrasound, a contrast enhanced pancreatic CT, or an MRCP, a panoply of opportunities, some that either create a drainage tract, remove a stone, or create an additional pathway. The management concerns in intervening are the respiratory failure risk, the risk of deterioration, how frequently you need to monitor them, and this all occurs within the assessment of the degree of pancreatic inflammation and or necrosis severity. Part of what you need to do is mitigate risk, and there are two real risks that you try and mitigate. One is acute kidney injury by avoiding hypovolemia, and the other is abdominal compartment syndrome by avoiding the promotion of ascites from excess plasma volume expansion. Analgesia adequacy helps support respiratory function and may decrease the risk of acute respiratory failure. Since pancreatitis is painful, there is generally abdominal distension, and you really need people to engage in pulmonary hygiene so they don't have in-specific secretions, and hypoxic pulmonary vasoconstriction and RV strain. So you think about the impact of intra-abdominal hypertension. This is one of the better done studies. Not a lot of patients, mostly male, intra-abdominal hypertension occurred in 36, 17% mortality, 80% of that occurred in the first 72 hours. Those that had abdominal compartment syndrome had a much, much higher mortality. What you're looking at in gray are patients without intra-abdominal hypertension versus, in black, those with intra-abdominal hypertension. And note that that occurs pretty early. It can occur as early as six hours, but it's persistent. So there's certainly a time frame for you to evaluate and intervene. It complicates everything. That same metric, no intra-abdominal hypertension versus those with it. Apache 2 scores are higher in hospital days 1, 2, and 3. Capillary leak index, which supports studies, but also extravascular lung water accumulation, also higher. Overall, higher fluid balance, which aids in the creation of secondary compartment syndrome, and they tend to have higher SOPHIS scores. It's all a problem. So therefore, early assessment of intra-abdominal hypertension and therapeutic interventions designed to mitigate that risk become very important early on. And it's not something that you should wait for 48 hours to ask and answer that question. But in fact, in those that are admitted to your ICU with moderate or severe pancreatitis, you can ask that question, and you should ask that question on admission. Every single metric, mortality, necrosis, operative intervention, vasoactive needs, mechanical ventilation, bacteremia, it's all worse in those with hypertension. So what about ERCP? When do you get the endoscopist involved? In particular, it's great if they have gallstone pancreatitis and they have either obstruction or cholangitis. What you'd really like is for them to have cholangitis. Cholangitis generally has a decreased intensity of local and systemic complications, and draining their biliary tract often improves their survival in a way that the obstructed patient doesn't quite realize. And when do you do this? Certainly within 24 to 48 hours of their admission. Resuscitation before ERCP is often important unless they have concomitant bacteremia or they have a completely obstructed biliary tract. Now, if they have gallstone pancreatitis and they're a high-risk patient for having their gallbladder taken out to reduce recurrence because up to 40% just about will have recurrence if you don't take out their gallbladder, you can pursue sphincterotomy as part of their ERCP. If you do that within the first month, recurrence is much less likely. Some never occur, and it's less than 10% that do. So this is another option instead of laparoscopic cholecystectomy or open cholecystectomy during the resolving phase of gallstone pancreatitis. Let's talk about infection. If you look at, now this is a much larger collection of patients. You can see 4,100 here and 2,800 there. Sterile necrosis versus infected necrosis. Look at the mortality rates, 13% versus 28%. Infection correlates with mortality. This is not a surprise. This is part of the tenets of the surviving sepsis campaign to address infection as a driver of poor outcome. Since infection is really a driver of that outcome, should you use prophylactic antibiotics? When I was training, everyone had a prophylactic carbapenem. And yet, when you look at 14 randomized control trials, a little over 800 patients, more multidrug-resistant organisms, more pseudodifficile infection, and no benefit with regard to bacterial or fungal infections, no reduction in mortality, and no reduction in the absolute need for intervention either. So the answer for prophylactic antibiotics is no. When should intervention occur? True pancreatic necrosis, there really is minimal separation of devitalized tissue and a high solid-to-fluid ratio. There's transitional pancreatic necrosis, which is an evolution. There's organized pancreatic necrosis, clear separation of a fluid-filled cavity with a fibrous wall and a pseudocyst, a very well-developed wall. They each may have different timings. Indications for intervention in each of those is when you have infection. If the diagnosis of whether you have an infection is unclear, sampling is helpful. If you have a persistent biliary cause where you cannot clear the biliary tract and achieve drainage, or if there are concurrent complications such as organ space infection, fistula leak, or ischemia. So two interventions that are important to recognize is fine needle aspiration. It's generally not required. False negative rate is almost 1 in 5. It tends not to impact outcome unless you're not certain whether or not you have infection, in which case sampling a fluid collection to determine the presence of bacteria can be helpful. Surgery itself is very controversial. It used not to be. It tends to be resolved for those that have more than 50% of their pancreas involved. Those with deterioration and those with accelerating organ failure. Infected necrosis is a great driver of therapy. CAT scan identified gas collections and a positive fine needle aspiration should drive antibiotics. And some approach, a step-up approach which we'll talk about momentarily. CT guided drainage, which is helpful in a select few. Endoscopic transenteric drainage, which is great for pseudocysts that are closely applied. Sometimes they get infected and will reserve operation laparoscopic or open for last. This is a typical CAT scan appearance and you can see the dark areas. This is gas within this necrotic section of pancreas. This person deserves and merits intervention. Just like this person does. This is a more sharp image. You see how this is juxtaposed with the stomach. You might be able to do this with transenteric range or in fact transabdominal transgastric and intrapancreatic range, but it only drains liquid components. Some of the components here are semi-solid. They don't come out through your catheter. Therefore, the step-up approach has been identified and we'll talk about what that is. This is a study that compares the step-up approach. Minimally invasive versus open necrosectomy to major complications plus death 40% versus 68. You see in general a minimally invasive approach is better than taking them to the OR and exploring them in a very typical broad-based fashion. You can have a surgical step-up approach, but you can also have an endoscopic one. There are pros and cons. Surgical step-up approach. You can almost always get there and it's really effective, but there's a much higher risk of external pancreatic fistula and you need general anesthesia. The endoscopic approach tends to have less of an inflammatory response. You don't need to access the intra-abdominal wall. Sometimes that's very important for those that have had multiple prior procedures, but you may not be able to reach everything and you may need more than a couple of procedures and you must have an expert. What does this really mean? It means the following. So if you screen about 418 patients, 206 are infected, 158 are sterile, they pursue a conservative approach. If you divide them 51 endoscopic and 47 surgical, these are the ones that agreed, you can compare the surgical versus the endoscopic step-up approach. You can see that the risk of fistula much less reduced, 5% versus 32%. Going from drainage to necrosectomy endoscopically. They do pretty well. It takes longer if you do this surgically and days in hospital, shorter length of stay. This is operative necrosectomy. This is a transverse incision and this is my hand pulling the stomach and transverse colon up. This is another hand pulling the colon down and you can see this portion here. This is very bluntly being picked up. This is dead pancreas. On the other hand, you can through an endoscopic or laparoscopic approach place a deployable stent through the, and this is across mucosa. This is through the stomach. This is the endoscopic approach to get into that cavity. And you can use these kinds of probes that have wire mesh and you can debride these walls so that you can do this through an endoscope if things are closely applied from the stomach by deliberately placing that stent between the two of them. You can also do this through a laparoscopic approach accessing the retroperitoneum. If you're going to operate and you ask the question, when should I operate for infected necrosis? If you ask, should I operate 72 hours? All of these studies, now mind you, they're relatively old, but spans 1985 to 2010, fairly durably identify that you should wait. If you ask the question, how about if I wait 12 days? Is that better? You should wait. How about 30 days? Favors waiting longer than 30 days. So the surgical approach to this is something that you should keep in reserve, but you can use these other approaches much more readily. In fact, IR tends to do a lot of these in many places. So it doesn't matter whether it's being done endoscopically, being done surgically, using a laparoscopic approach, or interventional radiology-based stent-driven approaches. They all work better than open neck resectomy in terms of resolution and complications. When do you take out somebody's gallbladder? Finally, something for a surgeon to do. Before discharge, or two to four weeks after discharge. Before discharge is highly favored. You can think about whether or not they're medically unfit, and if they're medically unfit for cholecystectomy, ERCP and sphincterotomy is ideal. And should they have an ERCP at all? If there's no cholangitis, but a high probability of stones, that makes sense. Otherwise, you can proceed with your cholecystectomy, and if they have a problem with their LFTs thereafter, you can always get an MRCP and a post-op ERCP. Not everyone needs to have an ERCP duct clearance before they have their gallbladder out. So I'll leave you with these conclusions. Severe acute pancreatitis benefits from severity assessment, early enteral nutrition, a very prescribed role for ERCP to ask and answer specific questions related to obstruction, and whether or not there are stones in the duct. There is no role for prophylactic antibiotics, and a step-up approach endoscopically driven versus laparoscopically supported is favored, and you should hold surgery in reserve for everything other than cholecystectomy for those who have gallstone pancreatitis before they go home. Thank you very much. I appreciate your time and your attention. Thank you.

acute pancreatitis

etiologies

diagnostic tools

nutritional support

Revised Atlanta classification

interventions