E. Elton

Endoscopic pancreatic sphincterotomy: indications, outcome, and a safe stentless technique

BACKGROUND Endoscopic pancreatic sphincterotomy is less widely practiced than biliary sphincterotomy, in part because of the lack of firm data regarding its indications and safety. In addition, recent reports of ductal and parenchymal changes occurring after pancreatic stenting raise concerns about the standard practice of stent placement at the time of pancreatic sphincterotomy. We report our experience with pancreatic sphincterotomy and describe the use of a technique involving overnight nasopancreatic drainage rather than stenting. METHODS We reviewed the records of the 164 pancreatic sphincterotomies performed on 160 patients at our institution between January 1, 1991, and October 1, 1996, comparing procedures done with overnight nasopancreatic catheter placement with those done with stenting or no drainage. We also examined the long-term clinical outcome of patients after pancreatic sphincterotomy. RESULTS Of the 164 sphincterotomies, 98 were done with overnight nasopancreatic drainage, 50 with stent placement, and 16 with no drainage. Complications (all pancreatitis) were significantly more frequent in the group with no drainage (12.5%) as compared with those with drainage (0.7%); p < 0.003. Nasopancreatic drainage was as safe as stent placement, with no complications after 98 procedures. Pancreatic sphincterotomy was effective when used as primary therapy, with 64% of patients so treated experiencing complete and long-lasting resolution of symptoms after the procedure. CONCLUSIONS Pancreatic sphincterotomy is safe and effective, although pancreatic drainage is required to reduce the incidence of pancreatitis. Overnight nasopancreatic drainage is the method of choice, as it carries as low a complication rate as stent placement, but without the need for a repeat procedure, and presumably without the risk of ductal and parenchymal damage.

Diagnostic and therapeutic ERCP using an enteroscope and a pediatric colonoscope in long-limb surgical bypass patients

Diagnostic and therapeutic ERCP has nearly the same high success and low complication rates in Billroth II gastroenterostomy patients as in patients with normal anatomy.1,2 However, long-limb surgical bypasses have, in general, precluded endoscopic access for retrograde cannulation. Patients who have undergone these operations have usually been evaluated and treated with interventional radiologic procedures or surgery. With the advent and refinement of push enteroscopes and pediatric colonoscopes, however, endoscopic access to the papilla is now possible. We describe here our initial experience in using these endoscopes for retrograde cannulation and therapeutic ERCP in patients who have undergone long-limb bypasses.

Resolution of mediastinal pancreatic pseudocysts with transpapillary stent placement

Pancreatic pseudocysts can occur as a complication of acute or chronic pancreatitis. They are usually located in the peripancreatic area, most commonly in the lesser sac. On rare occasion a pseudocyst can reach the mediastinum. Approximately 40 such cases have been reported.1 These pseudocysts are caused by rupture of the pancreatic duct posteriorly into the retroperitoneal space. The pancreatic fluid then tracks up through a diaphragmatic hiatus into the mediastinum, usually the posterior mediastinum.2-4 In one series of 24 patients, the commonest sites of entry into the mediastinum from the diaphragm were the openings for the esophagus and the aorta (21 out of 24 patients).5 Other less frequent sites are the inferior vena cava hiatus (case 2 of our report), foramen of Morgagni, and direct penetration of the diaphragm. Traditionally, pancreatic pseudocysts, whatever the site, have been treated initially with conservative management and surgery with internal drainage for those cases in which the pseudocyst does not resolve with conservative treatment alone.1-7 More recently, endoscopic transpapillary stent placement has been introduced as a therapeutic option for pancreatic fistulas and communicating pseudocysts in patients who do not respond to conservative treatment.8-11 In other cases, endoscopic transmural drainage or percutaneous drainage has been used. This is a report of 2 cases of mediastinal pancreatic pseudocysts that were successfully treated with transpapillary stent placement as primary therapy in 1 case and in conjunction with percutaneous drainage in the other.

4607 Nasopancreatic drainage at ercp: techniques and complications.

INTRODUCTION: We have described a safe technique of pull-type endoscopic pancreatic sphincterotomy (EPS) with overnight nasopancreatic drainage to lower the risk of post procedural pancreatitis, (Gastrointest Endosc 1998;47:240-9). This avoids the need for pancreatic duct stenting which requires x-ray follow-up, frequent repeat endoscopy for removal, and possible pancreatic ductal injury which may not be completely reversible. The technique differs from nasobiliary drainage in significant ways. PATIENTS & METHODS: From 1/91-10/99, 238 patients underwent EPS using the following technique: After pancreatic duct cannulation, a small guide wire was advanced out to the tail followed by a monofilament pulltype EPS. Biliary ES was performed either prior to EPS or occasionally immediately after. A 5Fr nasopancreatic drain (Wilson-Cook, Winston- Salem, NC) was maneuvered out to the area of the mid-PD. The end was brought out through the right nares using a nasal transfer tube. Benzoin was applied to the lip and face and the 5Fr catheter was trimmed to 25cm from the nares. Adhesive (Tegaderm, 3M Health Care, St. Paul, MN) is placed over the entire length of the catheter from the nose to the lateral aspect of the neck. The Tuohy-Borst adapter was secured to the drainage tubing and pinned to the patients hospital robe. Drainage was then placed to a urinary bag and kept below the level of the patient to insure gentle siphonage. RESULTS: There were 11/238 (4%) complications of EPS including mild pancreatitis (n=8), moderate bleeding (n=1), mild perforation (n=1). There was 1 instance of premature nasopancreatic drain removal by a patient and no recognized incidence of accidental internal dislodgment of the transpapillary catheter. The nasopancreatic drains were left in place for 20-72 hours (mean=24 hours). All drains could be pulled comfortably at the bedside without sedation or excessive patient discomfort. There were no complications attributed to obstruction or perforation by the catheters. IMPRESSION: Nasopancreatic drainage is a simple and quickly performed procedure which provides adequate drainage following pancreatic sphincterotomy and obviates the need for stent placement. Careful attention to details such as avoiding over-insertion of the catheter and taping it securely to the patients face optimizes drainage and prevents premature withdrawal. Concern regarding frequent dislodgment, poor patient tolerance, and patient discomfort during withdrawal are unfounded.

⁎3499 Therapeutic ercp in the setting of long limb surgical bypass: improved success and outcome of therapeutic procedures.

INTRODUCTION: We reported initial ERCP experience in 18 long-limb surgical bypasses (Gastrointest Endosc 1998;47: 62-7). Only 6 had intact papillas and only 3 required sphincterotomy (ES). We report additional experience. PATIENTS & METHODS: From 4/93-8/99, 47 pts (male 18, female 29, age 24-80) were referred for consideration of therapeutic ERCP with long-limb surgical bypasses: 8 with gastric bypass; 16 with Whipples, 11 with hepatojejunostomy; and 12 with post-gastrectomy Roux-en-Ys. Indications: pain (n=26), jaundice (n=17), recurrent or chronic pancreatitis (n=11), suspected CD stones (n=7), and anastomotic bleeding (n=1). An enteroscope was used in 13, a pediatric colonoscope in 17 and a duodenoscope in 17. RESULTS: The papilla or pancreaticobiliary anastomosis was reached in 41 of 47 (87%). The overtube was used with 5 and only in the setting of an intact stomach. 6 failures were due to inability to pass into the afferent limb. 2 failures were due to inability to locate the PD anastomoses. 20 patients had intact papillas (12 with Roux-en-Ys, 8 with gastric bypass). Cannulation failed in 4 of these 20 (20%) due to an inability to align the long front viewing instruments. 7 pts with intact papillas needed ES; 3 methods were used: (1) Following guidewire and temporary stent placement, a stent guided needle-knife ES was used when alignment could be achieved (n=4); (2) A wire-guided B-II type sphinctertome was used (n=1). (3) One patient failing guidewire placement underwent successful needle knife ES only. The final patient required a combined procedure with PTC guidewire placement followed by B-II type ES at a second ERCP. Among the 35 cannulated patients, successful Rx followed in 34: stents in 16 (10 biliary plastic stents, 2 PD stents, 4 Wallstents); biliary dilatation in 6; stone extraction 3; tissue sampling 3; pancreatic dilatation 1; duodenal dilatation 1. In 1 patient, coagulation of a visible vessel required surgical treatment for recurrence. No post-ERCP pancreatitis, bleeding, or perforation was noted. CONCLUSION: Although technically difficult, successful therapeutic ERCP is generally possible with long limb surgical bypasses with a high success rate and a low complication rate. The intact papilla greatly exacerbates these difficulties and occasionally requires percutaneous wire placement for assistance. The development of a colonoscope length, oblique-viewing enteroscope with an elevator should greatly enhance therapeutic ERCP in these patients.

4602 Endoscopic pancreatic sphincterotomy: long-term follow-up.

INTRODUCTION: Endoscopic biliary sphincterotomy has just enjoyed its 25th anniversary, but endoscopic pancreatic sphincterotomy (EPS) has been performed for less than 10 years with little long-term follow-up data. We employ only pull-type EPS, and have described a safe, stentless technique (Gastrointestinal Endosc 1998;47:240-9). We report our long-term results. PATIENTS & METHODS: From 1/91 to 10/99, we performed 397 EPS procedures on 349 patients. There were 144 males, 205 females, age 5 to 90. The indications for sphincterotomy were SOD, recurrent acute pancreatitis, chronic pancreatitis, pancreatic duct stones, ampullary polypectomy, pancreatic fistula, and pseudocyst drainage. 238 EPS (60%) had nasopancreatic drainage, 132 (33%) were accompanied by PD stent placement, and 27 (7%) had no drainage. 332 (84%) EPS were of the major papilla in 293 patients and 65 (16%) were of the minor papilla performed in 56 patients. 10 patients had both major and minor duct sphincterotomies on different occasions. Follow-up has been for a median of 34 months with a minimum follow-up of 6 months. RESULTS: There were 22/397 (5%) shortterm complications consisting of mild pancreatitis (N=13), mild or moderate bleeding (N=2), mild perforation (N=2), cholangitis, stent migration, pain relieved by stent removal, and premature patient removal of a nasopancreatic drain (N=1 each). There were no requirements for surgery and no deaths. Long-Term Follow-Up: 39 of 397 (10%) EPS were repeat sphincterotomies performed in 34 patients for incomplete initial EPS (n=4), for additional therapies, such as stone removal, ampullary polypectomy or pseudocyst drainage (n=19), and for restenosis with recurrent symptoms (n=16). In these last 16 (4%), 6 were minor duct EPS (6/65=9.2%) compared to 10 major duct EPS (10/332=3%; p=0.03). Of the 16, 7 had underlying chronic pancreatitis characterized by ongoing disease and inflammation. There was only 1 complication in the repeat EPS group comprising minor bleeding after EPS extension to remove a migrated PD stent. IMPRESSION: Our previously described technique of pull-type pancreatic sphincterotomy is both safe and effective in moderate to long-term follow-up.We employ nasopancreatic drainage to avoid the risks and costs of pancreatic stenting when post sphincterotomy drainage beyond 24 hours is not anticipated. There is a very low incidence of restenosis. Minor duct sphincterotomy is more likely to restricture compared to major duct EPS.