Acute Pancreatitis: Necrosis, Vascular Traps, and Recurrence

0:00Welcome to Board Pearls. This is episode three of three of the Acute Pancreatitis chapter, in the Pancreatic and Biliary Disease module. This episode is the complications and special cases: necrotizing disease and its step-up management, the vascular complications, the prevention of post-ERCP pancreatitis, and the workup of the patient who keeps coming back.

0:22The episode picks up where the pancreas has declared itself necrotic. About one in four or five patients develops necrotizing disease, and the rule for spotting it is single: on a contrast CT the pancreatic parenchyma fails to enhance, because the gland has lost perfusion. Interstitial edematous pancreatitis enhances, necrotizing pancreatitis does not, and that one distinction gates the rest of the episode.

0:47Most necrosis is sterile in the first week, and infection emerges between weeks two and four, and it doesn't come from the bloodstream, it comes from the gut, across the same wall whose integrity we protected with early feeding in the last episode, when the patient is sick enough and the tissue beyond it is dead enough. Most infections in acute pancreatitis are in necrotizing disease, and a substantial fraction of necrosis ultimately gets infected, and that matters because sterile necrosis carries roughly a ten percent mortality while infected necrosis runs much higher, which is why recognizing infection is urgent and why one CT finding is essentially diagnostic.

1:24That finding is gas in the collection: loculated air inside a peripancreatic collection, absent recent intervention, means gas-forming gut organisms are in there and can't be read as anything else, and the clinical correlate is new or persistent fever, rising white count, and deterioration after an initial improvement. Endoscopic-ultrasound-guided aspiration with gram stain and culture exists but isn't the routine move, because antibiotics are usually started empirically on clinical suspicion plus imaging, and aspiration is for the equivocal case where the decision actually turns on microbiologic confirmation.

2:02The nomenclature for the collections is where tested vocabulary lives, and it organizes on two things: timing and content. Timing draws a line at four weeks, and content draws a line between fluid alone and fluid containing necrotic tissue. Before four weeks without necrosis is an acute peripancreatic fluid collection, and most resolve on their own. Before four weeks with necrosis is an acute necrotic collection. After four weeks without necrosis, encapsulated, is a pseudocyst. And after four weeks with necrosis, encapsulated, is walled-off necrosis. The favorite confusion is that last one, because an encapsulated collection more than four weeks after necrotizing pancreatitis with heterogeneous density and solid debris inside is walled-off necrosis, not a pseudocyst, and the name matters because a pseudocyst is fluid you can drain through a small conduit while walled-off necrosis contains solid tissue that will obstruct a small conduit, so recognizing the debris means recognizing you may need to debride, not just drain.

3:03The step-up principle is three words, delay, drain, debride, and each is a separate teaching. Delay, because intervention before four weeks carries higher mortality, since the necrosis hasn't demarcated, viable and dead tissue aren't cleanly separated, and early debridement disrupts living gland and worsens bleeding, fistula, and organ failure, so the four-week clock is the time it takes the body to wall off the dead tissue. Drain percutaneously or endoscopically first, because the less invasive option may resolve the sepsis with far lower morbidity, so source control without surgical injury is the goal. And debride only if drainage fails to clear the sepsis or the collection holds solid debris that obstructs drainage, because necrosectomy is the highest-morbidity intervention here and earns its place only when drainage can't get there.

3:53This sequence was established when infected necrotizing pancreatitis was randomized between primary open necrosectomy and a step-up approach starting with percutaneous drainage and escalating only if drainage failed, and the step-up arm substantially cut both death-or-major-complications and new multi-organ failure, because removing source control without disrupting viable tissue lets the patient stabilize, and the patient who stabilizes survives.

4:19The next question was whether the step-up itself goes better by surgical or endoscopic route, and while mortality was similar, the endoscopic route produced far fewer pancreatic fistulas and shorter stays, and the mechanism is anatomic: the endoscopic transmural route enters the collection through the stomach or duodenum and never crosses the peritoneum, so there's no surgical wound to fistulate, which is why endoscopic drainage with a lumen-apposing metal stent, and direct necrosectomy through it when needed, has become the modern standard where it can be done.

4:47A few situations force intervention before four weeks: hemodynamic instability from infected necrosis with documented gas, persistent organ failure refractory to medical management, or a fistula that won't close. In any of these the preferred move is percutaneous catheter drainage rather than surgery, because it temporizes and buys time to get the patient stable enough that definitive debridement can wait until the necrosis has demarcated and the operation is safer. And two points on antibiotics: prophylactic antibiotics for sterile necrosis aren't recommended, because there's no benefit and they drive resistance and C. difficile, while for infected necrosis or strong suspicion the agents are carbapenems, fluoroquinolones, or piperacillin-tazobactam with metronidazole because those penetrate pancreatic tissue, with cultures then guiding narrowing.

5:36That handles necrosis. The vascular complications are mechanistically distinct from the inflammatory core and each has its own trigger. On the venous side, splanchnic thrombosis develops when peripancreatic inflammation damages the wall of the splenic, portal, or superior mesenteric vein, and most of these clots resolve on their own over weeks to months, so the instinct to anticoagulate everyone with a clot on the scan is wrong here, because anticoagulation adds bleeding risk in a patient with necrosis and the natural history is favorable without it, and you reserve it for clot extending into the portal or superior mesenteric vein where mesenteric ischemia or hepatic decompensation is the concern, while isolated splenic vein thrombosis without extension doesn't need systemic anticoagulation.

6:15That isolated splenic vein thrombosis has a chronic counterpart worth naming, left-sided portal hypertension, whose signature is isolated gastric varices in a patient with normal liver function, and whose treatment if those varices bleed is splenectomy, a lesion that recurs in the chronic pancreatitis chapter.

6:33The arterial side is more dangerous: a pseudoaneurysm of the gastroduodenal, splenic, or hepatic artery is the bleed that kills, formed when activated enzymes leaking into peripancreatic tissue erode the arterial wall directly, the surrounding fibrosis briefly contains the defect, and then it ruptures massively. It presents as GI bleeding, sudden enlargement of a known collection, or an unexplained hemoglobin drop in a patient with known necrosis or a pseudocyst, and pancreas-protocol CT diagnoses it while angiography is both diagnostic and therapeutic.

7:09This is where the high-yield procedural rule lives: before any necrosectomy or cyst drainage in a patient with a peripancreatic collection, image the arterial anatomy, and if a pseudoaneurysm is present, embolize it first, because pushing a stent or scope into a collection that hides an unrecognized pseudoaneurysm risks catastrophic hemorrhage when the manipulation disrupts the fibrous shell, so embolization first and drainage second is non-negotiable, and the same rule governs cyst drainage in chronic pancreatitis. So the necrotic patient going to the endoscopy suite for stent deployment gets a pancreas-protocol CT angiogram first, and the team looks at the arteries before the collection, because the collection is the target but the artery determines whether the target is safe to enter.

7:51That brings the episode to post-ERCP pancreatitis, which fits here because it's an etiology the GI hospitalist sees and because the prevention triad is high-yield. It hits three to fifteen percent of cases and is the most common ERCP complication, and the risk factors divide into patient, procedure, and operator, with the organizing logic that each pillar of prevention attacks a different node of the injury cascade, so combined prophylaxis beats any single intervention because the nodes are independent.

8:21The patient factors are what the boards lean on. The single highest is suspected sphincter of Oddi dysfunction, then prior post-ERCP pancreatitis, with female sex, younger age, and a normal bilirubin all raising risk, and the bilirubin point is the one most learners miss, so reason through it: a normal bilirubin means the indication isn't obstruction, which means the duct isn't dilated, which makes cannulation harder, with more wire passes into the pancreatic duct and more papillary trauma, so the patient with a high bilirubin and a dilated duct is the easy, low-risk cannulation and the patient with a normal bilirubin going for suspected sphincter dysfunction is the difficult, high-risk one, with the bilirubin acting as a proxy for cannulation difficulty. Chronic pancreatitis is paradoxically protective, because the fibrotic gland mounts less inflammatory response to the same insult, a board-tested counterintuitive finding.

9:17The procedure factors are direct mechanical insults: difficult cannulation, meaning more than ten minutes or five attempts, contrast injection into the pancreatic duct especially to the point of opacifying the acinar bed, pancreatic sphincterotomy, precut papillotomy, balloon dilation of an intact sphincter, and ampullectomy. And the operator factor reduces to volume, because high-volume operators cannulate with fewer attempts and fewer pancreatic duct injections, so the real risk factor is the number of wire passes the patient endures.

9:46The three pillars each act on a different mechanism. First, rectal indomethacin one hundred milligrams around the time of the procedure, where NSAID blockade of the inflammatory enzymes attenuates the cascade triggered by acinar injury and roughly halves the incidence, acting at the cytokine level and doing nothing about mechanics or perfusion. Second, a prophylactic pancreatic duct stent in high-risk cases, meaning difficult cannulation, suspected sphincter dysfunction, ampullectomy, prior post-ERCP pancreatitis, or pancreatic sphincterotomy, where the mechanism is mechanical, because the papilla swells after the procedure and transiently obstructs pancreatic outflow which drives autodigestion, and a small-caliber stent across the papilla keeps the duct patent through that edematous window, acting at the obstruction level and doing nothing about cytokines. Third, peri-procedural lactated Ringer's, aggressive fluid loading around the procedure that provides microcirculatory protection and inflammatory dilution, preferred over saline for the same acidosis-avoidance logic as the resuscitation choice in episode one, acting at the perfusion level and doing nothing about cytokines or obstruction. Because the three attack independent nodes, high-risk cases routinely receive all three, and no single one protects against the others' failure modes.

11:04The final piece is the patient who recovers, goes home, and returns with a second attack. Roughly a fifth of patients have a second attack within five years, and recurrent acute pancreatitis is a major risk factor for chronic pancreatitis, so finding the cause is preventive medicine for the next disease as much as the next attack. Idiopathic acute pancreatitis is not a presentation-day diagnosis, it's the label only after a full baseline workup is unrevealing, and that baseline is a careful alcohol and drug history, a lipid panel checked when the patient isn't acutely third-spaced, a calcium and an IgG4, and an ultrasound for stones, because most patients labeled idiopathic on a first pass never actually got a complete first pass, missing occult alcohol use, certain drugs, or a transient triglyceride spike that had normalized by the time it was measured.

11:55When the baseline is genuinely negative and the patient has a second attack, the second-pass workup is MRCP plus endoscopic ultrasound, with a meaningful yield of new findings that cluster into occult duct stones, pancreas divisum, a mucinous neoplasm, autoimmune pancreatitis, and early chronic changes.

12:14The discriminating point between the two is that endoscopic ultrasound is more sensitive for tumors and for microlithiasis, particularly over forty, so the patient with unexplained recurrent pancreatitis after forty and a clean MRCP gets endoscopic ultrasound, asking whether a small tumor or biliary microlithiasis was missed at lower resolution, and microlithiasis, the sludge and tiny stones that ultrasound and MRCP both miss, is the highest-yield finding in this group. Under thirty-five with recurrent attacks and a family history, the workup turns genetic, with the inherited-pancreatitis panel, and the dominant trypsinogen mutation carries a markedly elevated lifetime adenocarcinoma risk that drives surveillance.

12:56Pancreas divisum is the favorite anatomic trap, so reason through it carefully. In development the dorsal and ventral pancreatic ducts normally fuse, and when they fail to, the dorsal duct drains most of the gland through the small minor papilla while the ventral duct drains only the inferior head and uncinate through the major papilla, and the proposed mechanism for pancreatitis is relative outflow obstruction at a small papilla draining a large gland. The prevalence is in the range of five to fourteen percent, and that's the trap, because most people with divisum never get pancreatitis, so finding it in a patient with recurrent attacks doesn't establish it as the cause, it may be coincidental. When this was tested directly, randomizing recurrent-pancreatitis patients with divisum between minor papilla sphincterotomy and sham, sphincterotomy did not reduce future attacks, so it isn't recommended for divisum on the basis of an imaging finding alone, and it's reserved for symptomatic divisum with documented obstructive features like dorsal duct dilation, remaining a center-of-expertise procedure even then.

13:56The single highest-yield prevention is stopping the inciting cause. Alcohol cessation started during the index admission with counseling repeated as an outpatient substantially reduces recurrence, smoking cessation addresses an independent risk for both chronic pancreatitis and cancer, and lipid lowering for hypertriglyceridemia uses fibrates first-line with omega-threes and diet, targeting a triglyceride under five hundred. After a second idiopathic episode, guidance suggests cholecystectomy, which substantially reduces recurrence, the inferred mechanism being that occult biliary microlithiasis is the missed cause in a meaningful fraction and removing the gallbladder resolves it without ever identifying the specific stone, making cholecystectomy the high-yield empirical move when imaging can't find a target. And post-discharge surveillance addresses the metabolic sequelae, screening for new-onset diabetes with a hemoglobin A1c at three to six months and annually, since a substantial fraction develop it, and monitoring for exocrine insufficiency, which develops in up to a third at a year, watching for steatorrhea, weight loss, and fat-soluble vitamin deficiency, with the reason to monitor everyone being that even more than one in ten mild cases will have exocrine insufficiency at a year.

15:13Pull it together. Necrosis runs on the timing-and-content nomenclature and step-up management, delay then drain then debride, with endoscopic transmural drainage now favored because the route avoids the peritoneum and the fistulas that follow. The vascular complications split by vessel type, venous thrombosis usually self-resolving and pseudoaneurysm demanding pre-drainage CT angiography because the alternative is catastrophic. The prevention triad attacks three independent nodes, cytokines with indomethacin, obstruction with a duct stent, and perfusion with lactated Ringer's, so combined prophylaxis outperforms any single pillar. And recurrent or idiopathic disease is worked up by escalation, baseline labs then MRCP then endoscopic ultrasound for microlithiasis and tumors, with genetic testing added in the young patient and cholecystectomy the empirical move after a second idiopathic episode for unseen microlithiasis.

16:10Chapter twenty-six picks up the chronic disease the recurrent patient may be drifting toward: chronic pancreatitis and how it's diagnosed and its pain and enzyme replacement managed, the pancreatogenic diabetes, the autoimmune pancreatitis subtypes, the cystic lesions and their risk stratification, and pancreatic adenocarcinoma.

16:29For 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 three of three of chapter twenty-five, and I'll see you in the next one.