ERCP and EUS Procedures: Fluid Collections and Necrosectomy

0:00Welcome to Board Pearls. This is episode six of seven of the ERCP and EUS Procedures chapter, in the Endoscopic Procedures module. In this episode we cover the collections the gland leaves behind: the maturation timing that decides when to drain, the pseudoaneurysm check that prevents a catastrophe, the step-up from transmural drainage to direct endoscopic necrosectomy, and the carbon dioxide rule that keeps necrosectomy from causing an air embolism.

0:27The last episode stayed inside the gland, the recurrent attack and the chronic dilated duct. Now the spectrum shifts again, from the gland itself to the collections it leaves behind. A patient who had necrotizing pancreatitis weeks ago now has a symptomatic fluid collection. The first decision is timing, and it is the same wall-maturation logic throughout. You wait until about four weeks, because that is the time the inflammatory rind needs to mature into a wall strong enough to hold a transmural anastomosis without leaking. Drain too early and the wall cannot support the connection you are about to create.

1:06The second decision is what is actually in the cavity, because that drives everything downstream. A pseudocyst is homogeneous fluid with a mature wall after four weeks and no significant solid debris. Walled-off necrosis is heterogeneous, a mature-walled collection that contains solid necrotic debris, the classic sequel of necrotizing pancreatitis. That distinction is not academic. It determines the stent you choose and whether you will ever need to go back in and debride.

1:35Before you puncture anything, you review the contrast cross-sectional imaging, the CT angiogram, for a pseudoaneurysm. This is the step that prevents a catastrophe. Necrotizing pancreatitis erodes arterial walls with activated enzymes, so a pseudoaneurysm can sit in the cavity wall or in the splenic, gastroduodenal, or pancreaticoduodenal distribution. If you needle into a cavity that contains or shares a wall with a pseudoaneurysm, you can cause uncontrolled hemorrhage. So when a pseudoaneurysm is present, interventional radiology embolizes it first, and only then do you drain.

2:10The drainage strategy itself is a step-up, and the principle is to do the least invasive thing that works and escalate only on failure. The step-up approach was first established for surgery, and the endoscopic version used here, transmural drainage first and necrosectomy only when the response is inadequate, is supported by randomized trials. The reason transmural drainage leads is that the endoscopic route does not cross the peritoneum or retroperitoneum, so it avoids the surgical wound complications that drive so much of the morbidity in a surgical step-up.

2:42Stent caliber follows the contents directly. A simple pseudocyst with thin fluid drains fine through a single lumen-apposing metal stent or through two double-pigtail plastic stents. Thin contents flow easily through a small lumen, and plastic is a reasonable lower-cost option when a LAMS is not available. Walled-off necrosis is different, because solid debris cannot pass through a small lumen. So you need a large-caliber LAMS, the fifteen or twenty millimeter sizes, large enough that the endoscope itself can pass through the tract for direct necrosectomy if you need it. The ten millimeter LAMS is for the simple pseudocyst where you do not anticipate going back in. Sizing is a trade: a bigger lumen handles debris and reduces the need for later dilation, but every stent also carries a migration risk.

3:32Even after drainage, the step-up logic continues. You do not jump to necrosectomy because there is necrosis on the scan. You drain transmurally first and watch the clinical response. You only escalate to necrosectomy when the response is inadequate: persistent fever, persistent leukocytosis, persistent solid debris on imaging. The cavity tells you whether it is emptying.

3:56Dwell time has its own reasoning. After walled-off necrosis drainage you typically reassess and remove the LAMS at about four weeks, with four to eight weeks as the acceptable range. That timing is guided by clinical and imaging resolution rather than a fixed calendar date. Pull it too early and the tract collapses before the cavity has finished draining. Three recognized LAMS complications are worth holding in mind. Delayed bleeding, which is often pseudoaneurysm-related. Buried LAMS syndrome, where the stent erodes into the wall. And migration of the stent. Once the cavity has shrunk and the patient has resolved, you remove the LAMS and place double-pigtail plastic stents across the now-mature tract for continued internal drainage.

4:39Two failure modes deserve to be over-learned, because they test heavily and the wrong move is dangerous. The first is LAMS occlusion. The patient was draining fine, then returns with recurrent fever and rising inflammatory markers, and the cavity is not emptying because retained necrotic material has plugged the stent. The maneuver is urgent endoscopy to clear the LAMS lumen and place double-pigtail plastic stents through the LAMS. Those pigtails work because they hold a smaller patent channel open even as debris re-accumulates around them.

5:10The second is bleeding into the LAMS tract. A pseudoaneurysm ruptures into the cavity, and the patient presents with melena, hematemesis, or hemorrhagic shock coming through the existing stent. Here the sequence is everything. Interventional radiology embolizes the pseudoaneurysm before the LAMS comes out. The LAMS may be tamponading the bleeding source, so pulling it first, before you have controlled that source, releases the tamponade and produces uncontrolled hemorrhage. Embolize, then remove. That same rule, control the source before you release the stent, applies anywhere a LAMS tract is bleeding.

5:45There is one diagnostic parallel in this territory. Suppose a patient had pancreatitis and is left with a persistent tail cyst, and imaging cannot tell whether it communicates with the duct. The question is whether this is a benign pseudocyst or a mucinous neoplasm, and you answer it with EUS-guided aspiration and cyst fluid analysis. The numbers track epithelial behavior. A CEA above one hundred ninety-two favors mucinous, with good specificity but modest sensitivity. A glucose below fifty favors mucinous, and it is the more sensitive marker. The mechanism is clean and worth holding. Mucinous epithelium is metabolically active and consumes glucose, so the fluid runs low, while pseudocyst fluid simply retains glucose diffused from serum. High amylase tells you the cavity communicates with the duct, consistent with a pseudocyst or a branch-duct IPMN, and cytology adds the dysplasia read.

6:41That brings us to the cavity that will not clear on drainage alone. Direct endoscopic necrosectomy is the through-LAMS debridement you perform when transmural drainage has not cleared the solid debris in walled-off necrosis. It is the modern standard, having largely replaced surgical necrosectomy, for the same reason transmural drainage leads the step-up. The endoscopic route does not breach the peritoneum or retroperitoneum, so it avoids the fistulas, prolonged ileus, and wound infections that drive surgical morbidity.

7:12The technique follows from what the cavity is. You pass a forward-viewing therapy endoscope through the LAMS into the cavity and use snares, baskets, nets, and forceps to mechanically remove necrotic tissue. You do this across several sessions, typically two to four and sometimes up to six, and the reason is biological. The necrotic tissue is adherent to underlying granulation tissue, so trying to strip the whole cavity in one sitting tears that granulation bed and bleeds. You take a portion at a time and let the rest demarcate. Dilute three percent hydrogen peroxide irrigation is sometimes used to loosen and oxidize adherent debris, though the practice is not universal.

7:53The single non-negotiable safety rule is the insufflation gas, and the reason is anatomy. The cavity may communicate with vascular structures or with the peritoneum. If you insufflate with air, that gas can enter the splenic vein or portal venous radicles. It can also pass through a free communication. Either way it can cause an air embolism. So carbon dioxide insufflation is mandatory during necrosectomy. Carbon dioxide is rapidly absorbed, so even if it crosses into the venous system it produces no clinical embolic event. The same physics bans forceful flushing under pressure, because rapidly pressurizing the cavity drives gas or fluid into those same vascular communications. This is the scenario where the patient suddenly collapses with a falling end-tidal carbon dioxide and air in the right ventricle, and the lesson is to use carbon dioxide and to stop forcefully flushing. The recognized complications of necrosectomy are bleeding from granulation tissue or a wall pseudoaneurysm, perforation through the cavity wall, and infection from gut flora reseeding the cavity. The fourth is that air embolism. Between sessions the LAMS stays in for ongoing drainage, you decompress with a nasogastric tube if there is ileus, and you continue antibiotics in infected necrosis.

9:11Pull the collections together. You wait for a mature wall at about four weeks and you image for a pseudoaneurysm first, embolizing before you ever puncture. The contents decide the stent, a small lumen for thin pseudocyst fluid and a large-caliber LAMS for the solid debris of walled-off necrosis. Then you drain transmurally and escalate to direct endoscopic necrosectomy only when the cavity tells you it is not emptying, debriding a portion at a time across several sessions under carbon dioxide insufflation. The two dangerous failure modes both turn on sequence: clear an occluded LAMS with pigtails through it, and in a bleeding tract embolize the source before you release the stent. The cavity, the contents, and the timing decide the answer.

9:54In the next and final episode we turn EUS from a guide for therapy into a diagnostician and a palliator. We cover tissue acquisition, where the needle and the sonographic layer together decide what you can prove. And we cover celiac plexus neurolysis, where the same probe that finds the tumor is used to treat its pain.

10:14For the full chapter, the practice vignettes, and the topic-tagged question bank, head to board pearls dot com. You'll find the rest of the series on Apple Podcasts, Spotify, or wherever you listen to podcasts. That brings us to the end of episode six of seven of chapter twenty nine, and I'll see you in the next one.