Acute Pancreatitis: Acute Pancreatitis: Necrosis 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 necrosis and what comes after the attack. The first patient is the man from the last two episodes, febrile in week three, and his case shows when necrosis becomes infected and how to know, who needs antibiotics, how collections are named and why the name matters, why the order is delay, drain, then debride, and which vascular complications must be found before anyone drains. A woman with sterile necrosis shows who doesn't need antibiotics. The second patient is a young woman with recurrent attacks, and her case shows how to prevent the next one: when recurrent pancreatitis can be called idiopathic, who needs genetic testing, whether pancreas divisum explains anything, who's at risk from endoscopic retrograde cholangiopancreatography, ERCP, and how that risk is prevented, why the gallbladder may come out without stones, and what to check after discharge.

1:00Here's the first patient. He's the sixty-three-year-old man with severe gallstone pancreatitis and thirty percent necrosis on his day-four computed tomography, CT. His kidney failure recovered, and by day ten he was improving. In week three he develops a fever of thirty-nine and a rising white count, and he looks worse. His repeat CT shows a large collection in the pancreatic bed with pockets of gas. He's had no procedures. His surgeon wants to operate and remove the dead tissue now.

1:31Fever, a rising white count, and gas in a necrotic collection in week three mean infected pancreatic necrosis, but his surgeon's timing is wrong. About twenty to twenty-five percent of patients with acute pancreatitis develop necrotizing pancreatitis, recognized on contrast CT by gland that doesn't enhance because it isn't perfused. Most necrosis is sterile in the first week, and infection emerges in weeks two to four as gut bacteria cross the inflamed, ischemic gut wall into the dead tissue. About ninety percent of infections in acute pancreatitis occur in necrotizing disease, and twenty to seventy percent of necrosis becomes infected, which is why enteral feeding, which protects the gut barrier, matters.

2:12His diagnosis rests on his scan, and his scan is enough. Gas in a collection without a recent procedure is the most specific imaging finding and is pathognomonic for infected necrosis. The clinical signs are new or persistent fever, a rising white count, and deterioration after an initial improvement, and he has all of them. Aspiration guided by endoscopic ultrasound, EUS, with gram stain and culture is reserved for equivocal imaging, when the management decision depends on the microbiology, so his gas makes the diagnosis without it. Infected necrosis carries a mortality of fifteen to fifty percent, compared with about ten percent for sterile necrosis, which is why recognizing infection is urgent.

2:52His infection is treated first with antibiotics, chosen for their penetration. Infected or strongly suspected infected necrosis gets intravenous antibiotics that penetrate pancreatic tissue: a carbapenem, a fluoroquinolone, or piperacillin-tazobactam plus metronidazole, narrowed later by cultures from drainage or aspiration. He starts a carbapenem.

3:14The same drugs would harm a patient without infection. A forty-five-year-old woman with alcohol-related necrotizing pancreatitis and forty percent necrosis on day five has no fever, no rising white count, and no gas. She doesn't get prophylactic antibiotics, because they don't help sterile necrosis and they raise antibiotic resistance and the risk of Clostridioides difficile infection. She's watched and treated if she deteriorates.

3:40His collection has a name, and the name decides how it's drained. The Atlanta terms distinguish four collections by timing and content. An acute peripancreatic fluid collection is early, contains no necrosis, and usually resolves. A pseudocyst is encapsulated fluid without necrosis more than four weeks after onset. An acute necrotic collection contains necrosis with fluid in the first four weeks. Walled-off necrosis is encapsulated necrotic tissue after four weeks. His, with necrosis in week three, is an infected acute necrotic collection. An encapsulated collection after necrotizing pancreatitis with heterogeneous density and solid debris is walled-off necrosis, not a pseudocyst, whatever older term a report uses, because solid debris doesn't drain like fluid.

4:27Operating now would be the mistake, and the step-up order explains why. The sequence is delay, drain, debride. Delay about four weeks when possible, because intervention before the necrosis has demarcated disrupts viable tissue and raises mortality, bleeding, fistulas, and persistent organ failure. Drain first, percutaneously or endoscopically, because less invasive drainage has lower morbidity and may resolve the sepsis on its own. Debride only if drainage fails to control sepsis or solid debris blocks continued drainage, because necrosectomy is the highest-morbidity step.

5:05The trials support each part of that order. In a randomized trial of infected necrosis, percutaneous drainage first, with minimally invasive necrosectomy only if needed, substantially reduced death and major complications, and new multiple organ failure, compared with primary open necrosectomy. A second trial comparing endoscopic with surgical step-up found similar mortality but far fewer pancreatic fistulas and a shorter stay with the endoscopic approach, because going through the stomach wall crosses neither the peritoneum nor the retroperitoneum and creates no surgical wound to fistulate. Endoscopic step-up uses a lumen-apposing metal stent, LAMS, to create a cyst-gastrostomy, with direct endoscopic necrosectomy if needed, as chapter twenty-nine covers. He's stable on antibiotics with a collection that hasn't walled off, so the answer is not now.

5:55Earlier intervention is reserved for patients who can't wait. It's indicated for hemodynamic instability from infected necrosis with documented gas, persistent organ failure despite medical treatment, or a fistula, and even then percutaneous catheter drainage is preferred, as a bridge to debridement after the necrosis demarcates. A blood pressure of eighty-five in week two despite fluids and antibiotics would have meant a percutaneous catheter, not open necrosectomy. He stabilizes, and by week five his collection has walled off. Endoscopic drainage through his stomach with a LAMS settles his fever, and he doesn't need necrosectomy.

6:32Before anyone drains, vascular complications have to be excluded, and his blood count shows why. Before his drainage, his hemoglobin falls from eleven to eight over two days without obvious bleeding, and his collection looks larger. Pseudoaneurysms of the gastroduodenal, splenic, or hepatic artery form when activated enzymes erode the arterial wall, and they present with gastrointestinal bleeding, sudden enlargement of a collection, or an unexplained fall in hemoglobin. A pancreas-protocol CT diagnoses them, and angiography both diagnoses and treats them. Embolization is required before any drainage or necrosectomy, because manipulating a collection that contains a pseudoaneurysm risks catastrophic hemorrhage. His CT angiogram shows a splenic artery pseudoaneurysm, which is embolized first, with drainage two days later.

7:23His CT also shows a splenic vein thrombus, which needs a different decision. Splanchnic venous thrombosis, of the splenic, portal, or superior mesenteric vein, develops from inflammatory damage to the vein wall and mostly resolves over weeks to months. Anticoagulation is reserved for extension into the portal or superior mesenteric vein, because that's where mesenteric ischemia or hepatic decompensation become the risk. His clot is confined to the splenic vein, so he isn't anticoagulated. Its chronic counterpart, sinistral portal hypertension with isolated gastric varices, treated with splenectomy if they bleed, is watched for, as chapter twenty-six covers.

8:05The second patient's problem isn't one attack but the next one. A thirty-three-year-old woman is admitted with her second mild attack in fourteen months. She drinks rarely, doesn't smoke, and takes no medicines. Her triglycerides, calcium, and immunoglobulin G4 are normal, and two ultrasounds showed no gallstones. Her mother had pancreatitis twice in her forties. Her magnetic resonance cholangiopancreatography, MRCP, showed pancreas divisum, and her team has referred her for ERCP with minor papilla sphincterotomy.

8:38She has recurrent acute pancreatitis, and before it's called idiopathic, or blamed on her divisum, the causes have to be worked through. Roughly seventeen to twenty-three percent of patients have a second attack within five years, and recurrent acute pancreatitis is a major risk factor for chronic pancreatitis. Idiopathic pancreatitis means recurrent attacks without a cause after a baseline workup of alcohol and drug history, lipids, calcium, immunoglobulin G4, and ultrasound, which she's completed. A second-pass workup with MRCP and EUS finds a cause in thirty to sixty percent: occult duct stones, pancreas divisum, intraductal papillary mucinous neoplasm, autoimmune pancreatitis, and early chronic pancreatitis. EUS is more sensitive than MRCP for small tumors and microlithiasis in patients over forty, so it's the test for an older patient with an unexplained attack. Her EUS shows no microlithiasis, mass, or chronic pancreatitis, and confirms her divisum.

9:39Her age and her mother's attacks call for genetic testing. Testing for PRSS1, SPINK1, CFTR, CTRC, and CASR is reasonable in patients under thirty-five with recurrent attacks, particularly with a family history, and she meets every part of that. PRSS1 hereditary pancreatitis, autosomal dominant and highly penetrant, would fit her mother's history, and it carries a fifty-fold lifetime risk of adenocarcinoma that drives surveillance, as chapter twenty-six covers along with total pancreatectomy with islet autotransplantation.

10:16Her divisum probably isn't the cause, even though it's the finding that prompted her referral. Pancreas divisum is the failure of the dorsal and ventral ducts to fuse, so the dorsal duct, of Santorini, drains the body and tail through the minor papilla, and the ventral duct, of Wirsung, drains only the inferior head and uncinate process through the major papilla. The proposed mechanism is relative outflow obstruction through a small minor papilla draining most of the gland. But divisum is present in five to fourteen percent of people at autopsy, and most people with it never get pancreatitis, so finding it doesn't establish cause.

10:51The trial evidence settles her referral. A sham-controlled trial published in twenty twenty-five, in recurrent acute pancreatitis with divisum, found that minor papilla sphincterotomy didn't reduce attacks. So it's not recommended on the strength of the imaging finding alone, and it's reserved for symptomatic divisum with documented obstructive features, a dilated dorsal duct or focal pancreatitis at the minor papilla, at an expert center. She has neither feature, so her ERCP is cancelled, sparing her a procedure that wouldn't help.

11:22Her ERCP would also have carried a real risk of post-ERCP pancreatitis. It follows three to fifteen percent of ERCPs and is the most common complication. Risk is highest with suspected sphincter of Oddi dysfunction, a prior episode of post-ERCP pancreatitis, a young woman with a normal bilirubin, difficult cannulation, and pancreatic duct injection or sphincterotomy, while chronic pancreatitis is paradoxically protective. She'd have carried several of those risks. When an ERCP is truly needed in a high-risk patient, three measures act on different steps of the injury: rectal indomethacin, one hundred milligrams, which roughly halves post-ERCP pancreatitis in high-risk patients; a prophylactic pancreatic duct stent, which keeps the duct draining while the papilla swells; and lactated Ringer's solution around the procedure. The combination works better than any one of them, as chapter twenty-nine covers, and once post-ERCP pancreatitis occurs, it's managed like any other attack.

12:26Her genetic panel and her gallbladder decide what happens next. Her panel shows a SPINK1 variant, a permissive modifier that lowers her threshold for pancreatitis rather than causing it on its own, which makes avoiding alcohol and tobacco even more important for her. After a second idiopathic attack, the twenty twenty-four American College of Gastroenterology guideline suggests cholecystectomy, because in pooled data recurrent attacks were far less common after it, with occult biliary microlithiasis the inferred missed cause. With nothing else explaining her attacks, cholecystectomy is discussed with her.

13:03When the cause is known, removing it is the highest-yield prevention. Alcohol cessation starts during the first admission with formal counseling and continues as an outpatient, and it substantially reduces recurrence. Smoking cessation goes with it, because smoking is an independent risk factor for chronic pancreatitis and pancreatic cancer. The thirty-nine-year-old man from episode one with triglyceride-induced pancreatitis lowers his triglycerides below five hundred milligrams per deciliter, with fibrates first line, omega-3 fatty acids and diet as adjuncts, and treatment of his newly found diabetes.

13:39After discharge, every patient needs follow-up for what the attack may have damaged. New-onset diabetes is screened with a hemoglobin A1c at three to six months and then yearly. It occurs in up to twenty-five percent of patients in retrospective studies, and in seven to ten percent at one year in prospective studies. Exocrine insufficiency develops in about a quarter to a third at one year, more often with male sex, alcohol as the cause, greater severity, and necrosis, but in more than ten percent of mild cases too, so every patient is asked about steatorrhea, weight loss, and fat-soluble vitamin deficiency. Enzyme dosing and type three c diabetes are covered in chapter twenty-six. The man with necrotizing pancreatitis carries the most risk factors.

14:26Here are the traps the boards set in necrosis. Sterile necrosis doesn't get prophylactic antibiotics. Gas in a collection proves infection without an aspirate. Walled-off necrosis isn't a pseudocyst. Infected necrosis isn't operated on in the first weeks, and open necrosectomy doesn't come before drainage. The endoscopic approach causes fewer fistulas. A falling hemoglobin means excluding a pseudoaneurysm before any drainage. And not every splenic vein thrombosis is anticoagulated.

14:56For recurrence, the traps are these. Recurrent pancreatitis isn't called idiopathic without MRCP or EUS. EUS finds microlithiasis and small tumors in patients over forty. A young patient with recurrent attacks and a family history is offered genetic testing. Divisum on imaging isn't proof of cause, and the minor papilla isn't cut without obstructive features. A high-risk ERCP needs indomethacin, a pancreatic stent, and lactated Ringer's. Cholecystectomy is considered after a second idiopathic attack. The triglyceride target is below five hundred. And diabetes and exocrine insufficiency are screened for after discharge.

15:38So, start to finish. Name collections by content and time: fluid only is an acute peripancreatic collection, then a pseudocyst after four weeks, and necrosis is an acute necrotic collection, then walled-off necrosis after four weeks. Don't give prophylactic antibiotics for sterile necrosis. Suspect infection with gas or deterioration in weeks two to four, and give antibiotics that penetrate necrosis. Delay about four weeks, drain endoscopically or percutaneously first, and debride only if drainage fails. Embolize a pseudoaneurysm before any drainage, and anticoagulate splanchnic thrombosis only with portal or mesenteric extension.

16:21Before calling recurrent pancreatitis idiopathic, do a second-pass MRCP and EUS. Test genes in patients under thirty-five with recurrent attacks, especially with a family history. Don't cut the minor papilla for divisum seen on imaging alone. Prevent post-ERCP pancreatitis in high-risk patients with indomethacin, a pancreatic stent, and lactated Ringer's. Remove the cause, consider cholecystectomy after a second idiopathic attack, and screen for diabetes and exocrine insufficiency.

16:55That completes the Acute Pancreatitis chapter. Chapter twenty-six is chronic pancreatitis, cystic lesions, and pancreatic neoplasms.

17:05For 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.