- Email: [email protected]
Prospective, randomized, controlled trial of prophylactic pancreatic stent placement for endoscopic snare excision of the duodenal ampulla
ORIGINAL ARTICLE
Prospective, randomized, controlled trial of prophylactic pancreatic stent placement for endoscopic snare excision of the duodenal ampulla Gavin C. Harewood, MD, MSc, Nicole L. Pochron, LPN, Christopher J. Gostout, MD Rochester, Minnesota, USA
Background: Tumors that arise in the region of the major duodenal papilla account for 5% of GI neoplasms and 36% of resectable pancreaticoduodenal tumors. There is limited published literature that addresses the safety of endoscopic excision of the papilla. Although there is consensus about prophylactic pancreatic-duct stent placement, there is little supporting prospective data. The aim of this randomized, controlled trial was to compare the rates of postsnare ampullectomy pancreatitis in patients who did/did not receive prophylactic pancreatic-duct stent placement. Methods: Consecutive patients who were to undergo en bloc snare ampullectomy were randomized to placement of pancreatic-duct stent after ampullectomy or to no stent placement. Results: In total, 19 patients were enrolled, and 10 received pancreatic stents. Postprocedure pancreatitis occurred in 3 patients in the 24 hours after endoscopy, all cases occurred in the unstented group, 33% vs. 0% (stented group), p Z 0.02. Median peak amylase level was 3692 U/L (range 1819-4700 U/L) and median peak lipase level was 11450 U/L (range 5900-17,000 U/L). All 3 patients were hospitalized for a median of 2 days (range 1-6), and all made a complete recovery. Conclusions: Our findings suggest that a protective effect is conferred by pancreatic stent placement in reducing postampullectomy pancreatitis. Future large-scale studies are required to confirm this benefit. (Gastrointest Endosc 2005;62:367-70.)
Tumors that arise in the region of the major duodenal papilla account for 5% of GI neoplasms and 36% of resectable pancreaticoduodenal tumors.1 Adenoma is a particularly common finding in patients with familial adenomatous polyposis (FAP), occurring between 50% and 100%.2 Less commonly, an adenoma that involves the papilla occurs sporadically in patients without FAP. As with adenomatous polyps elsewhere within the GI tract, papillary adenomas are premalignant. Therefore, removal of adenomas at this site appears justified. Snare removal of the entire papilla as a single specimen was advocated as early as 1993.3 However, there is a paucity of published information regarding the safety of excision of the papilla.3,4 Postprocedure pancreatitis has been reported and may be greatest within the first 6 to 12 hours after the procedure, leading some investigators to propose prophylactic placement of a pancreatic-duct stent
after excision.3,5 There is consensus but very little prospectively acquired data that supports prophylactic pancreatic-duct stent placement. There is no data as to how long the duct should be stented. The aim of this prospective, randomized, controlled trial was to compare the rates of postprocedure pancreatitis in patients who undergo en bloc snare ampullectomy who do/do not receive prophylactic pancreatic-duct stent placement. We hypothesized that stent placement would reduce the rate of postprocedure pancreatitis.
PATIENTS AND METHODS Patients
Copyright ª 2005 by the American Society for Gastrointestinal Endoscopy 0016-5107/$30.00 doi:10.1016/j.gie.2005.04.020
Patients referred between January 2001 and May 2004 for endoscopic snare excision of an ampullary adenoma were considered for the study. Ampullary adenomas were defined as those confined on and/or within the ampulla of Vater. The study was approved by our institutional review board, and written consent was obtained from all patients who enrolled. After enrollment, patients, who were
www.mosby.com/gie
Volume 62, No. 3 : 2005 GASTROINTESTINAL ENDOSCOPY 367
Pancreatic duct stenting for snare ampullectomy
blinded to their assigned study arm, were randomized to one of two study groups by a study coordinator (N.L.P.) by using card selection from a sealed, opaque envelope: placement of pancreatic-duct stent after ampullectomy or no stent placement.
Harewood et al
Capsule Summary What is already known on this topic d
Endoscopic management All procedures were performed by a single endoscopist (C.J.G.), who has had experience in performing more than 50 prior endoscopic snare ampullectomies. All procedures were preceded by retrograde cholangiopancreatography to exclude intraductal extension of the lesion. Snare excision was performed for lesions confined to the papilla. Excision was performed with either a small snare (Bard Interventional Products, Billerica, Mass) or a prototype fine-wire, flexible snare designed for excision of the papilla (Olympus America Corp, Inc, Melville, NY). All procedures were performed by using blended monopolar current at a setting of ‘‘blend 2’’ on the electrosurgical generator (Meditron, Hackensack, NJ), which was equivalent to 25% of the duty cycle, with the power set at 52 W or else the ERBE generator (ERBE USA Inc, Marietta, Ga) was used at a standard Endocut sphincterotomy power setting of 250 W. A biliary sphincterotomy was performed to aid future surveillance and to prevent stenosis of the biliary orifice. A pancreatic-duct stent was inserted immediately after excision in those patients assigned to stent placement. A short, straight, single flanged, 3- or 5-cm-long, 5F polyethylene stent was used (Wilson Cook Medical Inc, Winston-Salem, NC) to facilitate spontaneous migration. At 24 hours after placement, fluoroscopy was performed and a plain abdominal radiograph was obtained. The stent, if still present, was removed endoscopically.
Patient follow-up Procedures were performed on an outpatient basis, but patients were asked to remain in the vicinity for 24 hours after the procedure. Patients who reported postprocedure abdominal pain were evaluated, and serum amylase and lipase levels were obtained. Pancreatitis was defined by a 3-fold elevation of pancreatic enzymes in the context of abdominal pain.6 All patients were contacted 24 hours after the procedure by a study coordinator, who was blinded to the patient’s assigned study arm, to ascertain any postprocedure symptoms.
En bloc snare ampullectomy for tumors of the duodenal ampulla is increasingly utilized but its safety is of paramount concern.
What this study adds to our knowledge d
In a single-center, prospective, randomized controlled trial, pancreatic stent placement reduced post-ampullectomy pancreatitis.
interim analysis of the results was planned if pancreatitis rates reached 30% in either study arm, because this was considered excessively high. After enrollment of 19 patients, the pancreatitis rate in the unstented arm had reached 30%, which prompted an interim analysis of the findings. After this analysis, which demonstrated marked differences in pancreatitis rates between the two study arms, our institutional review board expressed reservations about proceeding, and the decision was made to close the study to further enrollment.
RESULTS Baseline features In total, 19 consecutive patients were considered eligible for this study, and all patients consented to participate. Patients were randomized to pancreatic-duct stent placement (11 patients) and no stent placement (8 patients). One patient, who was assigned to the stent arm became, paradoxically, agitated during the procedure, which precluded stent placement. Therefore, final stent placement was performed in 10 patients. The two patient groups were well matched with respect to age and gender, as illustrated in Table 1. Of those patients with information available, ampullectomy was performed by using the ERBE generator in 11 of 13 patients.
Patient outcome
Outcomes in patients with and without prophylactic placement of a pancreatic-duct stent were compared. Proportions were compared by using the chi-square test, whereas continuous variables were compared with the Wilcoxon rank sum test for nonparametric data. We initially had intended to enroll 25 patients in each study arm, which would have geared the study with 80% power to detect a 25% difference in rates of pancreatitis between both groups, with alpha Z 0.05. A priori, an
Postprocedure pancreatitis occurred in 3 patients in the 24 hours after endoscopy; all cases occurred in the group without pancreatic-duct stents (33% vs. 0% [stented group], p Z 0.02). Median peak amylase level was 3692 U/L (range 1819-4700 U/L) (normal: 26-102 U/L) and median peak lipase level was 11,450 U/L (range 590017,000 U/L) (10-73 U/L). All 3 patients were hospitalized for a median of 2 days (range 1-6 days), and all made a complete recovery. Another patient in the unstented group was hospitalized for observation of abdominal pain. This resolved overnight and was not accompanied by elevations in amylase and lipase. Of the patients who
368 GASTROINTESTINAL ENDOSCOPY Volume 62, No. 3 : 2005
www.mosby.com/gie
Statistical analysis
Harewood et al
experienced pancreatitis, all had the ampullectomy performed with the ERBE generator. Of note, the patient in whom agitation precluded stent placement did develop postprocedure pancreatitis. Excluding this patient from the analysis still maintained similar rates of pancreatitis, 25% (unstented) vs. 0% (stented), p Z 0.05. By intentionto-treat analysis (i.e., including this patient in the stented arm), clinically appreciable differences persisted, 25% (unstented) vs. 9% (stented), although statistical significance disappeared (p Z 0.3). All 10 patients who underwent pancreatic-duct stent placement underwent fluoroscopy and had a plain abdominal radiograph 24 hours after the procedure. The stent was still in situ in 4 patients and required endoscopic removal with a snare. One patient who had a pancreatic-duct stent placed complained of nausea without abdominal pain when contacted 24 hours after the procedure. Cholestatic liver enzyme abnormalities were detected: alkaline phosphatase 1120 U/L (37-98 U/L), bilirubin 4.9 mg/dL (0.1-1.0 mg/dL), aspartate aminotransferase 171 U/L (12-31 U/L), alanine aminiotransferase 140 U/L (9-29 U/L). Pancreatic enzymes remained normal. This was attributed to biliary edema surrounding the biliary orifice even though a biliary sphincterotomy had been performed. Therefore, a biliary stent (along with the previously placed pancreatic-duct stent) was placed to improve biliary drainage.
Pancreatic duct stenting for snare ampullectomy
TABLE 1. Baseline features of patients who underwent snare ampullectomy With PD stent, n Z 10
Without PD stent, nZ9
p Value
Age, median (range), y
53.5 (29-76)
44 (30-65)
0.53
Gender, n (% men)
5 (50%)
2 (22%)
0.21
Ampullectomy current device (Meditron/ERBE)*
1/6
Postprocedure pancreatitis
0
1/5
3 (33%)
0.9
0.02
PD, Pancreatic duct. *Of those patients with information available.
This is the first prospective, randomized, controlled study to assess the impact of prophylactic pancreatic-duct placement on the occurrence of pancreatitis after snare ampullectomy. Our findings suggest that prophylactic pancreatic-duct stent placement has a beneficial effect in reducing rates of postampullectomy pancreatitis. However, further larger-scale prospective studies are required to confirm this protective effect. The duodenum, and particularly the peripapillary region, is the most common upper-GI site of malignancy in patients after colectomy, with FAP occurring in up to 4.5% to 8.5%.7,8 The risk of adenocarcinoma of the major papilla in patients with FAP has been estimated at greater than 100 times that of the general population.9 Several studies indicate that the median age at onset of peripapillary malignancy complicating FAP is in the 6th decade of life.7-9 Adenoma involving the papilla also is being increasingly recognized in the absence of FAP, often as an incidental finding at upper endoscopy. The traditional management of periampullary adenomas is surgical. Pancreaticoduodenectomy has been advocated,10 but the advantage of wide surgical clearance must be balanced against considerable morbidity and mortality.11,12 A less morbid procedure is transduodenal excision11-15; how-
ever, recurrence has been reported in 25% to 100% of patients after this procedure.11,13-16 The ideal endoscopic therapy for periampullary adenomas is uncertain. Endoscopic ablative therapy with monopolar fulguration was first published in 1989.17 Neodymium-yttrium aluminum garnet (Nd-YAG) laser also has been reported in a small series.18 However, in our experience, laser ablation led to deep-tissue injury, resulting in postpolypectomy syndrome and duodenal scarring. Thus, we have abandoned the Nd-YAG laser in favor of other modalities with relatively shallow injury (e.g., argon plasma coagulation and bipolar cautery). Snare ampullectomy is a newer technique that seems to be well tolerated.3,5,19,20 There is a paucity of studies in the medical literature that addresses the safety of endoscopic en bloc snare ampullectomy. Norton et al5 observed immediate postprocedure complications in 7 of 26 patients who underwent endoscopic ampullectomy, of whom 4 (15%) developed mild pancreatitis, two developed focal bleeding that was controlled by epinephrine injection, and one suffered a small retroduodenal perforation at the apex of the sphincterotomy incision. Interestingly, pancreatic stents had been placed in 10 patients, of whom two developed pancreatitis, whereas two of 18 without stents developed pancreatitis (p Z 0.5). In a series of 25 patients, Binmoeller et al3 reported immediate complications in 5 patients, 3 (12%) developed pancreatitis, and two developed minor bleeding (treated with epinephrine injection, without recurrence). Two smaller series have been reported in abstract form alone. In a series of 8 patients, Greenspan et al20 noted one episode of cholangitis after snare papillectomy but no other complications. Martin et al19 performed snare excision of the papilla in 14 patients; one required surgery for
www.mosby.com/gie
Volume 62, No. 3 : 2005 GASTROINTESTINAL ENDOSCOPY 369
DISCUSSION
Pancreatic duct stenting for snare ampullectomy
hemorrhage, and another died because of necrotizing pancreatitis. All snare ampullectomies were performed specifically without the creation of a submucosal fluid cushion (SFC) so that the deeper sphincteric musculature would be included in the resection, thereby permitting ready access to both the pancreatic-duct orifice and the common bile duct orifice. The use of an SFC converts en bloc ampullectomy to a mucosectomy, leaving an intact sphincter of Oddi and the need for formal cannulation of this undisturbed anatomy to regain access to both the pancreatic and common bile ducts. It has been our experience that this later situation may lend itself to a greater risk of postprocedure pancreatitis from excessive manipulation from postmucosectomy cannulation attempts. Prophylactic pancreatic stent placement before en bloc snare ampullectomy as performed in this series is not technically feasible. In cases when there is laterally spreading adenomatous tissue off the papilla, then the creation of a SFC in these regions, followed by piecemeal mucosal resection is desirable and safer for the thin-walled duodenum away from the ampulla. In cases when there are scant margins of suspicious polyp tissue along the ampullectomy resection line, these can be directly ablated by using the argon plasma coagulator. In our practice, this is performed with a power setting of 50 W and a gas flow of 1.0 L/min after the pancreatic stent is in place. The ERBE electrosurgical generator was used at the standard default setting used in our unit for the majority of the cases, because of the transition of our practice to the exclusive use of this device in the ERCP suites. The blended ‘‘Endocut’’ current of this device was considered desirable. Pure coagulation and pure-cutting currents can be used for en bloc snare ampullectomy. The presumption is that pure coagulation may increase the risk for postprocedure pancreatitis because of excessive and prolonged thermal injury, with resultant edema of the ampullary bed, and that pure-cut will result in greater risk of hemorrhage. In summary, our findings suggest that prophylactic pancreatic-duct stent placement confers a protective effect after en bloc snare ampullectomy by reducing postprocedure pancreatitis. However, further larger scale prospective studies are required to confirm these findings.
Harewood et al 2. Bleau BL, Gostout CJ. Endoscopic treatment of ampullary adenomas in familial adenomatous polyposis. J Clin Gastroenterol 1996;22:237-41. 3. Binmoeller KF, Boaventura S, Ramsperger K, Soehendra N. Endoscopic snare excision of benign adenomas of the papilla of Vater [see comment]. Gastrointest Endosc 1993;39:127-31. 4. Park SW, Song SY, Chung JB, Lee SK, Moon YM, Kang JK, et al. Endoscopic snare resection for tumors of the ampulla of Vater. Yonsei Med J 2000;41:213-8. 5. Norton ID, Gostout CJ, Baron TH, Geller A, Petersen BT, Wiersema MJ. Safety and outcome of endoscopic snare excision of the major duodenal papilla. Gastrointest Endosc 2002;56:239-43. 6. Cotton PB, Lehman G, Vennes J, Geenen JE, Russel RC, Meyers WC, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991;37:383-93. 7. Arvanitis ML, Jagelman DG, Fazio VW, Lavery IC, McGannon E. Mortality in patients with familial adenomatous polyposis. Dis Colon Rectum 1990;33:639-42. 8. Jagelman DG, DeCosse JJ, Bussey HJ. Upper gastrointestinal cancer in familial adenomatous polyposis. Lancet 1988;I:1149-51. 9. Offerhaus GJ, Giardiello FM, Krush AJ, Booker SV, Tersmette AC, Kelley NC, et al. The risk of upper gastrointestinal cancer in familial adenomatous polyposis [see comment]. Gastroenterology 1992;102: 1980-2. 10. Gray G, Browder W. Villous tumors of the ampulla of Vater: local resection versus pancreatoduodenectomy. South Med J 1989;82: 917-20. 11. Knox RA, Kingston RD. Carcinoma of the ampulla of Vater. Br J Surg 1986;73:72-3. 12. Cahen DL, Fockens P, de Wit LT, Offerhaus GJ, Obertop H, Gouma DJ. Local resection or pancreaticoduodenectomy for villous adenoma of the ampulla of Vater diagnosed before operation. Br J Surg 1997;84: 948-51. 13. Alstrup N, Burcharth F, Hauge C, Horn T. Transduodenal excision of tumours of the ampulla of Vater. Eur J Surg 1996;162:961-7. 14. Asbun HJ, Rossi RL, Munson JL. Local resection for ampullary tumors. Is there a place for it? Arch Surg 1993;128:515-20. 15. Farouk M, Niotis M, Branum GD, Cotton PB, Meyers WC. Indications for and the technique of local resection of tumors of the papilla of Vater. Arch Surg 1991;126:650-2. 16. Penna C, Phillips RK, Tiret E, Spigelman AD. Surgical polypectomy of duodenal adenomas in familial adenomatous polyposis: experience of two European centres. Br J Surg 1993;80:1027-9. 17. Shemesh E, Nass S, Czerniak A. Endoscopic sphincterotomy and endoscopic fulguration in the management of adenoma of the papilla of Vater. Surg Gynecol Obstet 1989;169:445-8. 18. Lambert R, Ponchon T, Chavaillon A, Berger F. Laser treatment of tumors of the papilla of Vater. Endoscopy 1988;20(Suppl 1):227-31. 19. Martin JA, Haber GB, Kortan PP, Raijman I, Abedi M, DuVall JA, et al. Endoscopic snare ampullectomy for resection of benign ampullary neoplasms [abstract]. Gastrointest Endosc 1997;45:AB139. 20. Greenspan AB, Walden DT, Aliperti G. Endoscopic management of ampullary adenomas [abstract]. Gastrointest Endosc 1997;45:AB133.
Received February 2, 2005. Accepted April 15, 2005.
REFERENCES 1. Scarpa A, Capelli P, Zamboni G, Oda T, Mukai K, Bonetti F, et al. Neoplasia of the ampulla of Vater. Ki-ras and p53 mutations. Am J Pathol 1993;142:1163-72.
370 GASTROINTESTINAL ENDOSCOPY Volume 62, No. 3 : 2005
Current affiliation: Division of Gastroenterology and Hepatology and Radiology, Developmental Endoscopy Unit, Mayo Clinic, Rochester, Minnesota, USA. Reprint requests: Christopher J. Gostout, MD, Division of Gastroenterology and Hepatology, Charlton 8, Mayo Clinic, Rochester MN 55905.
www.mosby.com/gie
Prospective, randomized, controlled trial of prophylactic pancreatic stent placement for endoscopic snare excision of the duodenal ampulla Gavin C. Harewood, MD, MSc, Nicole L. Pochron, LPN, Christopher J. Gostout, MD Rochester, Minnesota, USA
Background: Tumors that arise in the region of the major duodenal papilla account for 5% of GI neoplasms and 36% of resectable pancreaticoduodenal tumors. There is limited published literature that addresses the safety of endoscopic excision of the papilla. Although there is consensus about prophylactic pancreatic-duct stent placement, there is little supporting prospective data. The aim of this randomized, controlled trial was to compare the rates of postsnare ampullectomy pancreatitis in patients who did/did not receive prophylactic pancreatic-duct stent placement. Methods: Consecutive patients who were to undergo en bloc snare ampullectomy were randomized to placement of pancreatic-duct stent after ampullectomy or to no stent placement. Results: In total, 19 patients were enrolled, and 10 received pancreatic stents. Postprocedure pancreatitis occurred in 3 patients in the 24 hours after endoscopy, all cases occurred in the unstented group, 33% vs. 0% (stented group), p Z 0.02. Median peak amylase level was 3692 U/L (range 1819-4700 U/L) and median peak lipase level was 11450 U/L (range 5900-17,000 U/L). All 3 patients were hospitalized for a median of 2 days (range 1-6), and all made a complete recovery. Conclusions: Our findings suggest that a protective effect is conferred by pancreatic stent placement in reducing postampullectomy pancreatitis. Future large-scale studies are required to confirm this benefit. (Gastrointest Endosc 2005;62:367-70.)
Tumors that arise in the region of the major duodenal papilla account for 5% of GI neoplasms and 36% of resectable pancreaticoduodenal tumors.1 Adenoma is a particularly common finding in patients with familial adenomatous polyposis (FAP), occurring between 50% and 100%.2 Less commonly, an adenoma that involves the papilla occurs sporadically in patients without FAP. As with adenomatous polyps elsewhere within the GI tract, papillary adenomas are premalignant. Therefore, removal of adenomas at this site appears justified. Snare removal of the entire papilla as a single specimen was advocated as early as 1993.3 However, there is a paucity of published information regarding the safety of excision of the papilla.3,4 Postprocedure pancreatitis has been reported and may be greatest within the first 6 to 12 hours after the procedure, leading some investigators to propose prophylactic placement of a pancreatic-duct stent
after excision.3,5 There is consensus but very little prospectively acquired data that supports prophylactic pancreatic-duct stent placement. There is no data as to how long the duct should be stented. The aim of this prospective, randomized, controlled trial was to compare the rates of postprocedure pancreatitis in patients who undergo en bloc snare ampullectomy who do/do not receive prophylactic pancreatic-duct stent placement. We hypothesized that stent placement would reduce the rate of postprocedure pancreatitis.
PATIENTS AND METHODS Patients
Copyright ª 2005 by the American Society for Gastrointestinal Endoscopy 0016-5107/$30.00 doi:10.1016/j.gie.2005.04.020
Patients referred between January 2001 and May 2004 for endoscopic snare excision of an ampullary adenoma were considered for the study. Ampullary adenomas were defined as those confined on and/or within the ampulla of Vater. The study was approved by our institutional review board, and written consent was obtained from all patients who enrolled. After enrollment, patients, who were
www.mosby.com/gie
Volume 62, No. 3 : 2005 GASTROINTESTINAL ENDOSCOPY 367
Pancreatic duct stenting for snare ampullectomy
blinded to their assigned study arm, were randomized to one of two study groups by a study coordinator (N.L.P.) by using card selection from a sealed, opaque envelope: placement of pancreatic-duct stent after ampullectomy or no stent placement.
Harewood et al
Capsule Summary What is already known on this topic d
Endoscopic management All procedures were performed by a single endoscopist (C.J.G.), who has had experience in performing more than 50 prior endoscopic snare ampullectomies. All procedures were preceded by retrograde cholangiopancreatography to exclude intraductal extension of the lesion. Snare excision was performed for lesions confined to the papilla. Excision was performed with either a small snare (Bard Interventional Products, Billerica, Mass) or a prototype fine-wire, flexible snare designed for excision of the papilla (Olympus America Corp, Inc, Melville, NY). All procedures were performed by using blended monopolar current at a setting of ‘‘blend 2’’ on the electrosurgical generator (Meditron, Hackensack, NJ), which was equivalent to 25% of the duty cycle, with the power set at 52 W or else the ERBE generator (ERBE USA Inc, Marietta, Ga) was used at a standard Endocut sphincterotomy power setting of 250 W. A biliary sphincterotomy was performed to aid future surveillance and to prevent stenosis of the biliary orifice. A pancreatic-duct stent was inserted immediately after excision in those patients assigned to stent placement. A short, straight, single flanged, 3- or 5-cm-long, 5F polyethylene stent was used (Wilson Cook Medical Inc, Winston-Salem, NC) to facilitate spontaneous migration. At 24 hours after placement, fluoroscopy was performed and a plain abdominal radiograph was obtained. The stent, if still present, was removed endoscopically.
Patient follow-up Procedures were performed on an outpatient basis, but patients were asked to remain in the vicinity for 24 hours after the procedure. Patients who reported postprocedure abdominal pain were evaluated, and serum amylase and lipase levels were obtained. Pancreatitis was defined by a 3-fold elevation of pancreatic enzymes in the context of abdominal pain.6 All patients were contacted 24 hours after the procedure by a study coordinator, who was blinded to the patient’s assigned study arm, to ascertain any postprocedure symptoms.
En bloc snare ampullectomy for tumors of the duodenal ampulla is increasingly utilized but its safety is of paramount concern.
What this study adds to our knowledge d
In a single-center, prospective, randomized controlled trial, pancreatic stent placement reduced post-ampullectomy pancreatitis.
interim analysis of the results was planned if pancreatitis rates reached 30% in either study arm, because this was considered excessively high. After enrollment of 19 patients, the pancreatitis rate in the unstented arm had reached 30%, which prompted an interim analysis of the findings. After this analysis, which demonstrated marked differences in pancreatitis rates between the two study arms, our institutional review board expressed reservations about proceeding, and the decision was made to close the study to further enrollment.
RESULTS Baseline features In total, 19 consecutive patients were considered eligible for this study, and all patients consented to participate. Patients were randomized to pancreatic-duct stent placement (11 patients) and no stent placement (8 patients). One patient, who was assigned to the stent arm became, paradoxically, agitated during the procedure, which precluded stent placement. Therefore, final stent placement was performed in 10 patients. The two patient groups were well matched with respect to age and gender, as illustrated in Table 1. Of those patients with information available, ampullectomy was performed by using the ERBE generator in 11 of 13 patients.
Patient outcome
Outcomes in patients with and without prophylactic placement of a pancreatic-duct stent were compared. Proportions were compared by using the chi-square test, whereas continuous variables were compared with the Wilcoxon rank sum test for nonparametric data. We initially had intended to enroll 25 patients in each study arm, which would have geared the study with 80% power to detect a 25% difference in rates of pancreatitis between both groups, with alpha Z 0.05. A priori, an
Postprocedure pancreatitis occurred in 3 patients in the 24 hours after endoscopy; all cases occurred in the group without pancreatic-duct stents (33% vs. 0% [stented group], p Z 0.02). Median peak amylase level was 3692 U/L (range 1819-4700 U/L) (normal: 26-102 U/L) and median peak lipase level was 11,450 U/L (range 590017,000 U/L) (10-73 U/L). All 3 patients were hospitalized for a median of 2 days (range 1-6 days), and all made a complete recovery. Another patient in the unstented group was hospitalized for observation of abdominal pain. This resolved overnight and was not accompanied by elevations in amylase and lipase. Of the patients who
368 GASTROINTESTINAL ENDOSCOPY Volume 62, No. 3 : 2005
www.mosby.com/gie
Statistical analysis
Harewood et al
experienced pancreatitis, all had the ampullectomy performed with the ERBE generator. Of note, the patient in whom agitation precluded stent placement did develop postprocedure pancreatitis. Excluding this patient from the analysis still maintained similar rates of pancreatitis, 25% (unstented) vs. 0% (stented), p Z 0.05. By intentionto-treat analysis (i.e., including this patient in the stented arm), clinically appreciable differences persisted, 25% (unstented) vs. 9% (stented), although statistical significance disappeared (p Z 0.3). All 10 patients who underwent pancreatic-duct stent placement underwent fluoroscopy and had a plain abdominal radiograph 24 hours after the procedure. The stent was still in situ in 4 patients and required endoscopic removal with a snare. One patient who had a pancreatic-duct stent placed complained of nausea without abdominal pain when contacted 24 hours after the procedure. Cholestatic liver enzyme abnormalities were detected: alkaline phosphatase 1120 U/L (37-98 U/L), bilirubin 4.9 mg/dL (0.1-1.0 mg/dL), aspartate aminotransferase 171 U/L (12-31 U/L), alanine aminiotransferase 140 U/L (9-29 U/L). Pancreatic enzymes remained normal. This was attributed to biliary edema surrounding the biliary orifice even though a biliary sphincterotomy had been performed. Therefore, a biliary stent (along with the previously placed pancreatic-duct stent) was placed to improve biliary drainage.
Pancreatic duct stenting for snare ampullectomy
TABLE 1. Baseline features of patients who underwent snare ampullectomy With PD stent, n Z 10
Without PD stent, nZ9
p Value
Age, median (range), y
53.5 (29-76)
44 (30-65)
0.53
Gender, n (% men)
5 (50%)
2 (22%)
0.21
Ampullectomy current device (Meditron/ERBE)*
1/6
Postprocedure pancreatitis
0
1/5
3 (33%)
0.9
0.02
PD, Pancreatic duct. *Of those patients with information available.
This is the first prospective, randomized, controlled study to assess the impact of prophylactic pancreatic-duct placement on the occurrence of pancreatitis after snare ampullectomy. Our findings suggest that prophylactic pancreatic-duct stent placement has a beneficial effect in reducing rates of postampullectomy pancreatitis. However, further larger-scale prospective studies are required to confirm this protective effect. The duodenum, and particularly the peripapillary region, is the most common upper-GI site of malignancy in patients after colectomy, with FAP occurring in up to 4.5% to 8.5%.7,8 The risk of adenocarcinoma of the major papilla in patients with FAP has been estimated at greater than 100 times that of the general population.9 Several studies indicate that the median age at onset of peripapillary malignancy complicating FAP is in the 6th decade of life.7-9 Adenoma involving the papilla also is being increasingly recognized in the absence of FAP, often as an incidental finding at upper endoscopy. The traditional management of periampullary adenomas is surgical. Pancreaticoduodenectomy has been advocated,10 but the advantage of wide surgical clearance must be balanced against considerable morbidity and mortality.11,12 A less morbid procedure is transduodenal excision11-15; how-
ever, recurrence has been reported in 25% to 100% of patients after this procedure.11,13-16 The ideal endoscopic therapy for periampullary adenomas is uncertain. Endoscopic ablative therapy with monopolar fulguration was first published in 1989.17 Neodymium-yttrium aluminum garnet (Nd-YAG) laser also has been reported in a small series.18 However, in our experience, laser ablation led to deep-tissue injury, resulting in postpolypectomy syndrome and duodenal scarring. Thus, we have abandoned the Nd-YAG laser in favor of other modalities with relatively shallow injury (e.g., argon plasma coagulation and bipolar cautery). Snare ampullectomy is a newer technique that seems to be well tolerated.3,5,19,20 There is a paucity of studies in the medical literature that addresses the safety of endoscopic en bloc snare ampullectomy. Norton et al5 observed immediate postprocedure complications in 7 of 26 patients who underwent endoscopic ampullectomy, of whom 4 (15%) developed mild pancreatitis, two developed focal bleeding that was controlled by epinephrine injection, and one suffered a small retroduodenal perforation at the apex of the sphincterotomy incision. Interestingly, pancreatic stents had been placed in 10 patients, of whom two developed pancreatitis, whereas two of 18 without stents developed pancreatitis (p Z 0.5). In a series of 25 patients, Binmoeller et al3 reported immediate complications in 5 patients, 3 (12%) developed pancreatitis, and two developed minor bleeding (treated with epinephrine injection, without recurrence). Two smaller series have been reported in abstract form alone. In a series of 8 patients, Greenspan et al20 noted one episode of cholangitis after snare papillectomy but no other complications. Martin et al19 performed snare excision of the papilla in 14 patients; one required surgery for
www.mosby.com/gie
Volume 62, No. 3 : 2005 GASTROINTESTINAL ENDOSCOPY 369
DISCUSSION
Pancreatic duct stenting for snare ampullectomy
hemorrhage, and another died because of necrotizing pancreatitis. All snare ampullectomies were performed specifically without the creation of a submucosal fluid cushion (SFC) so that the deeper sphincteric musculature would be included in the resection, thereby permitting ready access to both the pancreatic-duct orifice and the common bile duct orifice. The use of an SFC converts en bloc ampullectomy to a mucosectomy, leaving an intact sphincter of Oddi and the need for formal cannulation of this undisturbed anatomy to regain access to both the pancreatic and common bile ducts. It has been our experience that this later situation may lend itself to a greater risk of postprocedure pancreatitis from excessive manipulation from postmucosectomy cannulation attempts. Prophylactic pancreatic stent placement before en bloc snare ampullectomy as performed in this series is not technically feasible. In cases when there is laterally spreading adenomatous tissue off the papilla, then the creation of a SFC in these regions, followed by piecemeal mucosal resection is desirable and safer for the thin-walled duodenum away from the ampulla. In cases when there are scant margins of suspicious polyp tissue along the ampullectomy resection line, these can be directly ablated by using the argon plasma coagulator. In our practice, this is performed with a power setting of 50 W and a gas flow of 1.0 L/min after the pancreatic stent is in place. The ERBE electrosurgical generator was used at the standard default setting used in our unit for the majority of the cases, because of the transition of our practice to the exclusive use of this device in the ERCP suites. The blended ‘‘Endocut’’ current of this device was considered desirable. Pure coagulation and pure-cutting currents can be used for en bloc snare ampullectomy. The presumption is that pure coagulation may increase the risk for postprocedure pancreatitis because of excessive and prolonged thermal injury, with resultant edema of the ampullary bed, and that pure-cut will result in greater risk of hemorrhage. In summary, our findings suggest that prophylactic pancreatic-duct stent placement confers a protective effect after en bloc snare ampullectomy by reducing postprocedure pancreatitis. However, further larger scale prospective studies are required to confirm these findings.
Harewood et al 2. Bleau BL, Gostout CJ. Endoscopic treatment of ampullary adenomas in familial adenomatous polyposis. J Clin Gastroenterol 1996;22:237-41. 3. Binmoeller KF, Boaventura S, Ramsperger K, Soehendra N. Endoscopic snare excision of benign adenomas of the papilla of Vater [see comment]. Gastrointest Endosc 1993;39:127-31. 4. Park SW, Song SY, Chung JB, Lee SK, Moon YM, Kang JK, et al. Endoscopic snare resection for tumors of the ampulla of Vater. Yonsei Med J 2000;41:213-8. 5. Norton ID, Gostout CJ, Baron TH, Geller A, Petersen BT, Wiersema MJ. Safety and outcome of endoscopic snare excision of the major duodenal papilla. Gastrointest Endosc 2002;56:239-43. 6. Cotton PB, Lehman G, Vennes J, Geenen JE, Russel RC, Meyers WC, et al. Endoscopic sphincterotomy complications and their management: an attempt at consensus. Gastrointest Endosc 1991;37:383-93. 7. Arvanitis ML, Jagelman DG, Fazio VW, Lavery IC, McGannon E. Mortality in patients with familial adenomatous polyposis. Dis Colon Rectum 1990;33:639-42. 8. Jagelman DG, DeCosse JJ, Bussey HJ. Upper gastrointestinal cancer in familial adenomatous polyposis. Lancet 1988;I:1149-51. 9. Offerhaus GJ, Giardiello FM, Krush AJ, Booker SV, Tersmette AC, Kelley NC, et al. The risk of upper gastrointestinal cancer in familial adenomatous polyposis [see comment]. Gastroenterology 1992;102: 1980-2. 10. Gray G, Browder W. Villous tumors of the ampulla of Vater: local resection versus pancreatoduodenectomy. South Med J 1989;82: 917-20. 11. Knox RA, Kingston RD. Carcinoma of the ampulla of Vater. Br J Surg 1986;73:72-3. 12. Cahen DL, Fockens P, de Wit LT, Offerhaus GJ, Obertop H, Gouma DJ. Local resection or pancreaticoduodenectomy for villous adenoma of the ampulla of Vater diagnosed before operation. Br J Surg 1997;84: 948-51. 13. Alstrup N, Burcharth F, Hauge C, Horn T. Transduodenal excision of tumours of the ampulla of Vater. Eur J Surg 1996;162:961-7. 14. Asbun HJ, Rossi RL, Munson JL. Local resection for ampullary tumors. Is there a place for it? Arch Surg 1993;128:515-20. 15. Farouk M, Niotis M, Branum GD, Cotton PB, Meyers WC. Indications for and the technique of local resection of tumors of the papilla of Vater. Arch Surg 1991;126:650-2. 16. Penna C, Phillips RK, Tiret E, Spigelman AD. Surgical polypectomy of duodenal adenomas in familial adenomatous polyposis: experience of two European centres. Br J Surg 1993;80:1027-9. 17. Shemesh E, Nass S, Czerniak A. Endoscopic sphincterotomy and endoscopic fulguration in the management of adenoma of the papilla of Vater. Surg Gynecol Obstet 1989;169:445-8. 18. Lambert R, Ponchon T, Chavaillon A, Berger F. Laser treatment of tumors of the papilla of Vater. Endoscopy 1988;20(Suppl 1):227-31. 19. Martin JA, Haber GB, Kortan PP, Raijman I, Abedi M, DuVall JA, et al. Endoscopic snare ampullectomy for resection of benign ampullary neoplasms [abstract]. Gastrointest Endosc 1997;45:AB139. 20. Greenspan AB, Walden DT, Aliperti G. Endoscopic management of ampullary adenomas [abstract]. Gastrointest Endosc 1997;45:AB133.
Received February 2, 2005. Accepted April 15, 2005.
REFERENCES 1. Scarpa A, Capelli P, Zamboni G, Oda T, Mukai K, Bonetti F, et al. Neoplasia of the ampulla of Vater. Ki-ras and p53 mutations. Am J Pathol 1993;142:1163-72.
370 GASTROINTESTINAL ENDOSCOPY Volume 62, No. 3 : 2005
Current affiliation: Division of Gastroenterology and Hepatology and Radiology, Developmental Endoscopy Unit, Mayo Clinic, Rochester, Minnesota, USA. Reprint requests: Christopher J. Gostout, MD, Division of Gastroenterology and Hepatology, Charlton 8, Mayo Clinic, Rochester MN 55905.
www.mosby.com/gie