ERCP and EUS Procedures: ERCP and EUS Procedures: Bile Duct and Ampulla

0:00Welcome to Board Pearls. This is episode two of seven of the ERCP and EUS Procedures chapter, in the Endoscopic Procedures module. This episode is proving whether a narrowed bile duct is cancer, and removing an ampullary adenoma without a Whipple. The first patient is a woman with sclerosing cholangitis and a new hilar stricture, and her case shows why brushings miss so many cancers, what forceps biopsy, fluorescence in situ hybridization, and cholangioscopy add, and why a needle into a hilar tumor is forbidden in a transplant candidate. A man with a pancreatic head mass shows when endoscopic ultrasound, EUS, is the better tool. A woman with pain after cholecystectomy shows when suspected sphincter of Oddi dysfunction is, and isn't, a reason for endoscopic retrograde cholangiopancreatography, ERCP. The second patient is a woman with an ampullary adenoma, and her case shows endoscopic ampullectomy: who qualifies, what has to be shown first, why it's done in one piece without a lift and with two stents, what goes wrong, and how she's followed. And a man with familial adenomatous polyposis shows why his duodenum is a different problem.

1:10Here's the first patient. A forty-seven-year-old woman has primary sclerosing cholangitis and ulcerative colitis. Over three months her alkaline phosphatase has doubled, she's become itchy, and her carbohydrate antigen 19-9 has risen to two hundred twenty. Her magnetic resonance cholangiopancreatography, MRCP, shows a new dominant hilar stricture involving the confluence of the right and left hepatic ducts. She's young and otherwise well, and her hepatologist is considering transplant. Two sets of ERCP brushings over the past month were negative.

1:44A new dominant hilar stricture with a rising carbohydrate antigen 19-9 in primary sclerosing cholangitis points to cholangiocarcinoma, and her negative brushings don't reassure. ERCP brush cytology alone has a sensitivity of only thirty to forty percent for malignancy in biliary strictures, because brushes sample only cells shed from the surface, and many cholangiocarcinomas are submucosal, infiltrating tumors that shed few cells into the duct. At that sensitivity, a negative result barely changes the probability, and her new stricture, her rising tumor marker, and her underlying disease still point to cancer.

2:23She needs better sampling, escalated step by step. Adding intraductal forceps biopsy and fluorescence in situ hybridization, FISH, raises sensitivity to about sixty percent. Forceps capture deeper tissue, through a cholangioscope or alongside a guidewire, and FISH detects polysomy, gains of several chromosomes, in shed cells that may be too sparse or bland for cytology. Cholangioscopy-directed biopsy raises sensitivity to more than seventy percent, because the endoscopist sees the duct lining and targets suspicious areas instead of sampling blindly. Its visual features, an irregular surface, tortuous tumor vessels, and friability, raise confidence even before histology, and mucosal pattern classifications refine the prediction. At her third ERCP, FISH shows polysomy, cholangioscopy shows an irregular, friable stricture with tortuous vessels, and targeted biopsy shows adenocarcinoma.

3:23Because she may be a transplant candidate, one way of sampling her is forbidden. In hilar cholangiocarcinoma, EUS-guided aspiration of the primary tumor is contraindicated in transplant candidates, because the needle tract crosses the peritoneal cavity and can seed tumor along it. That seeding creates extrahepatic disease, which disqualifies her from the Mayo Clinic transplant protocol, since the protocol requires the absence of extrahepatic disease. Percutaneous biopsy of the primary is avoided for the same reason. Brushings and intraductal biopsy at ERCP sample the primary without entering the peritoneum, which is why they're the preferred approach in transplant candidates.

4:06Suspicious lymph nodes are the exception to that rule, because sampling them can't create disease that isn't already there. Nodal disease is already extrahepatic spread, and positive nodes contraindicate transplant however they're found. So a suspicious regional node is sampled, by EUS if needed, and the procedure is framed for the patient as a test of transplant eligibility rather than diagnosis. Her nine-millimeter node near the hepatic artery is aspirated and is negative, so she's referred for the neoadjuvant transplant pathway, covered with the Bismuth-Corlette anatomy in chapter twenty-four.

4:42A distal stricture caused by a mass calls for EUS instead, because the mass itself can be sampled. A sixty-nine-year-old man has painless jaundice and a three-centimeter pancreatic head mass with a dilated duct ending abruptly at it. EUS-guided fine-needle aspiration or biopsy is better than ERCP brushings for a distal duct stricture or a pancreatic head mass, because EUS sees the mass directly and the needle samples its tissue rather than the duct surface. For pancreatic head adenocarcinoma, EUS fine-needle biopsy has a sensitivity above ninety percent. His adenocarcinoma is confirmed, giving the tissue needed before neoadjuvant chemotherapy, and his duct is drained at ERCP with a stent, as episode three covers.

5:30Tissue diagnosis is one reason for ERCP, and suspected sphincter of Oddi dysfunction is a reason that's often wrong. A forty-one-year-old woman had a cholecystectomy two years ago and has recurrent epigastric pain. Her liver tests are normal, and her common bile duct measures nine millimeters. She's been referred for ERCP with sphincter manometry. The Rome four classification and the sham-controlled trial that retired sphincterotomy for pain without objective abnormalities are covered in chapter twenty-seven, and the endoscopist's points follow from them. Sphincter of Oddi stenosis, the former type one, with pain, abnormal liver tests, and a dilated duct, gets biliary sphincterotomy without manometry. Suspected biliary sphincter dysfunction, the former type two, is an individualized, shared decision, and if ERCP is done, it's among the highest-risk procedures for post-ERCP pancreatitis, so rectal indomethacin, lactated Ringer's solution, and a prophylactic pancreatic duct stent are all used. Pain alone with normal tests and imaging, the former type three, isn't an indication, because sham did at least as well as sphincterotomy.

6:39Her referral falls in that last group. A duct of up to about ten millimeters after cholecystectomy with normal liver tests is physiologic, so her nine-millimeter duct isn't an abnormality, and her pain alone isn't an indication. Her referral is declined. Idiopathic recurrent pancreatitis without ductal disease likewise doesn't call for reflexive manometry or sphincterotomy.

7:05The second patient's lesion is at the ampulla itself, and the question is whether she needs a Whipple. A sixty-two-year-old woman had an esophagogastroduodenoscopy for dyspepsia showing an eighteen-millimeter, slightly raised, lobulated lesion at the major papilla, with tubular adenoma and low-grade dysplasia on biopsy. Her liver tests and lipase are normal, and she has no jaundice. Her surgeon has mentioned a Whipple.

7:31She has a sporadic ampullary adenoma with low-grade dysplasia, and she may not need a Whipple. Ampullary adenomas are found sporadically at endoscopy, or during duodenal surveillance in familial adenomatous polyposis, where the ampulla is the most common site of duodenal adenoma, and they progress through the adenoma-carcinoma sequence. That's why endoscopic ampullectomy is offered for benign or low-grade dysplastic lesions in patients who'd otherwise face pancreaticoduodenectomy. It trades the substantial morbidity of a Whipple for a procedure whose main complication, pancreatitis in ten to fifteen percent, is mostly managed conservatively or with repeat endoscopy.

8:13Before ampullectomy, her lesion has to be shown to be confined, because the snare can only remove what sits on the surface. EUS or intraductal ultrasound confirms there's no extension into the bile or pancreatic duct and no invasion, and cross-sectional imaging excludes liver metastases and lymph nodes. Intraductal extension contraindicates en bloc snare ampullectomy, because a snare can't capture tissue inside a duct, and residual disease drives recurrence, so a patient with intraductal extension goes to surgery. Hers is confined to the papilla. Biopsy understages about thirty percent of these lesions, with carcinoma found in the resected specimen despite a benign biopsy, so resection in one piece with full pathologic assessment is preferred over piecemeal removal.

8:56Her resection follows a set technique. Through a side-viewing duodenoscope, the lesion is captured in a snare and resected in one piece with electrocautery, for lesions up to two to three centimeters. Piecemeal resection is acceptable for larger lesions when one piece isn't feasible, accepting that piecemeal pathology can't stage as reliably. Submucosal injection is generally avoided, unlike with many colon polyps, because the submucosa at the ampulla is thin, and a lift pushes the lesion away from the snare. Hers is removed in one piece without injection.

9:31Her two ducts are then stented, because the resection bed swells. Both ducts are cannulated through the resection bed. A pancreatic duct stent, typically five French, prevents pancreatitis by keeping the duct draining through transient papillary edema. A biliary stent, typically seven to ten French plastic, prevents biliary obstruction from edema and eases access for later procedures. Both are removed at follow-up endoscopy four to eight weeks later. She also gets rectal indomethacin.

10:01Her complications are predictable from the procedure. Pancreatitis, at ten to fifteen percent, dominates. Bleeding occurs in two to eight percent, often delayed, and is treated at repeat endoscopy with clips or coagulation. Perforation occurs in less than five percent, often into the retroperitoneum, and a contained injury usually needs percutaneous drainage rather than surgery. An occluded biliary stent can cause cholangitis, and a stricture can form at the healing resection bed. Six days after her ampullectomy, she has melena with a two-gram hemoglobin fall, and a visible vessel is clipped at repeat endoscopy, a moderate adverse event under the lexicon in chapter twenty-eight, because it needed repeat endoscopic treatment. Fever and jaundice weeks later would point first to her biliary stent.

10:49She then needs follow-up, because recurrence is common. Her specimen shows tubular adenoma with low-grade dysplasia and clear margins, and her stents come out at six weeks. Surveillance endoscopy at three and six months detects local recurrence, which is commonly reported in about twenty to thirty percent overall, higher in familial adenomatous polyposis and after piecemeal resection, and often treatable endoscopically. At six months, a three-millimeter residual area at the edge of her scar is snared, and she stays in surveillance. Invasive carcinoma at the deep margin of her specimen would have meant surgery, with the ampullectomy having done its job as a staging procedure.

11:28The duodenum in familial adenomatous polyposis is a different problem, because the disease is diffuse. A thirty-four-year-old man with familial adenomatous polyposis had a colectomy at twenty and now has more than thirty duodenal adenomas, several over a centimeter, an enlarged adenomatous ampulla, and high-grade dysplasia in two biopsies. Duodenal surveillance in this disease follows the Spigelman criteria, which score polyp size, number, histology, and dysplasia, with side-viewing esophagogastroduodenoscopy every six to twelve months for established duodenal polyposis or a prior ampullary lesion. Spigelman stage four disease, with high-grade dysplasia or extensive ampullary involvement, may warrant pancreas-sparing duodenectomy or pancreaticoduodenectomy, because endoscopic management isn't durable when polyposis is diffuse. He's referred to discuss surgery. Familial adenomatous polyposis is covered in chapter thirty-six.

12:24Here are the traps the boards set in strictures and the sphincter. Negative brushings don't exclude cholangiocarcinoma, and sampling escalates to forceps biopsy, FISH, and cholangioscopy. A pancreatic head mass is sampled by EUS, not brushings. A hilar cholangiocarcinoma in a transplant candidate is never needled by EUS or percutaneously, but a suspicious node is sampled. Sphincter stenosis gets sphincterotomy without manometry. ERCP for suspected dysfunction needs full pancreatitis prophylaxis, and pain alone isn't an indication. And a post-cholecystectomy duct of up to ten millimeters with normal tests is normal.

13:06For the ampulla, the traps are these. A benign ampullary adenoma isn't sent straight to a Whipple. A lesion with intraductal extension isn't snared, and EUS or intraductal ultrasound comes first. A benign biopsy isn't final. Resection is done in one piece when possible, without a lift. A pancreatic stent is placed, and surveillance happens at three and six months. And diffuse duodenal polyposis in familial adenomatous polyposis isn't managed with repeated endoscopic resection.

13:38So, start to finish. Escalate stricture sampling: brushings detect thirty to forty percent of cancers, forceps biopsy with FISH about sixty, and cholangioscopy-directed biopsy more than seventy. Use EUS biopsy for a distal duct stricture or head mass, with sensitivity above ninety percent. In a possible transplant candidate with hilar cholangiocarcinoma, sample the primary at ERCP, never through the peritoneum, and do sample suspicious nodes. For the sphincter, give sphincterotomy without manometry for stenosis, make an individualized decision with full prophylaxis for suspected dysfunction, and don't do ERCP for pain alone.

14:18Offer ampullectomy for adenomas without intraductal extension or invasion on EUS or intraductal ultrasound, resect in one piece up to two to three centimeters without a lift, and place pancreatic and biliary stents for four to eight weeks. Expect pancreatitis in ten to fifteen percent, bleeding in two to eight, and perforation in less than five, and survey at three and six months for recurrence in a fifth to a third. In familial adenomatous polyposis, stage the duodenum by Spigelman and refer stage four disease for surgery.

14:49Episode three is malignant biliary obstruction: which stent for which patient, and EUS-guided drainage when ERCP fails.

14:58For 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 two of seven of chapter twenty-nine, and I'll see you in the next one.