Papillary tumors are unusual entities, making up less than 5% of all new digestive tract neoplasms in the United States. However, cancer of the pancreas is significantly more common, being responsible for approximately 12% of digestive organ cancers.
Periampullary cancers can be broadly considered as tumors arising within 1 cm of the ampulla of Vater and include ampullary, distal bile duct, pancreatic, and duodenal cancers. However, without careful histologic analysis, it is difficult, if not impossible, to differentiate the tumor type.
The accuracy of the preoperative diagnosis of ampullary carcinoma increased from 27% to 73% after the introduction of fiberoptic endoscopy. Malignant papillary tumors have a more favorable course after resection when compared with pancreatic cancer. When treated with the standard Whipple procedure, the 5-year survival rates in most series are approximately 35%. [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
A wide variation exists in the endoscopic appearance of normal major duodenal papillae. The only characteristic appearance distinguishing between benign and malignant disease is the presence of an ulcerative tumor mass, which has been observed in 21% of patients with malignancy and in no patients with benign disease.
Although malignant tumors of the main papilla are more common, a number of benign neoplasms also may arise in the periampullary area, including benign adenomas (tubular and villous), lipomas, hamartomas, fibromas, and neurogenic tumors. These last include neurofibromas, granular cell tumors, leiomyomas, lymphangiomas, and hemangiomas.
The most common benign tumor of the ampulla is the villous adenoma. The incidence is lower than that of ampullary carcinomas, being 0.04-0.12% in autopsy series. The average age at diagnosis is 62 years (range of 4mo to 79y).
In a number of cases, neuroendocrine tumors have both neurogenic and endocrine elements. Lesions of endocrine origin have been termed carcinoid tumors. Duodenal carcinoids have morphologic and functional similarities to pancreatic islet-cell tumors.
Neuroendocrine tumors are rare. Duodenal carcinoids account for only 1-5% of the total. Fewer than 40 cases of ampullary carcinoids exist in the world literature.
In 1987, the staging of ampullary carcinomas was added to the tumor, node, metastases (TNM) classification of malignant tumors by the International Union Against Cancer (now the Union for International Cancer Control), and since then, this classification has been used widely.
Of the 24 patients with carcinomas of the ampulla described by Mori et al, 5 had stage I disease (20.8%), 11 had stage II disease (45.8%), 7 had stage III disease (29.2 %), and 1 had stage IV disease (4.2 %) at diagnosis.
No specific laboratory studies exist for the diagnosis of pancreatic cancer, but the following symptoms may be seen:
Increased alkaline phosphatase, gamma–glutamyl transferase (GT), and direct bilirubin
Increased serum amylase and lipase
In a prospective, multicenter trial that included 30 patients with ampullary carcinoma, Bakkevold et al found that the sensitivities of several diagnostic studies in making a conclusive diagnosis were as follows:
Abdominal ultrasonography (23%)
Computed tomography (CT) scanning (58%)
Endoscopic retrograde cholangiopancreatography (ERCP, 78%)
ERCP and duct cytology (91%)
Percutaneous transhepatic cholangiography (100%)
In a series of 52 patients with biopsy-proven ampullary adenomas or carcinomas, Ponchon et al noted a normal endoscopic appearance of the papilla in 37% of patients. In addition, they found that 42% of intra-ampullary tumors became apparent only after endoscopic sphincterotomy. 
Grossly, papillary tumors may manifest as one of three well-described appearances, as follows:
Intramural protrusion - A bulge underneath a normal-appearing papilla
Exposed protrusion - Neoplastic-appearing tissue extending out from an abnormal-appearing papilla
In many cases, ascertaining the true source of these neoplasms is difficult. One beneficial aspect of these tumors, however, is that they frequently produce symptoms early in the course of the disease, when the tumor is relatively small. Most resection specimens contain tumors 1-2cm in size.
Almost all primary papillary cancers are adenocarcinomas, although sarcomas, lymphomas, carcinoid, and islet-cell cancers also have been reported. Histologically, papillary adenocarcinomas closely resemble adenocarcinomas of the colon. As in the colon, these neoplasms range from well differentiated to moderately differentiated to poorly differentiated. One series of 34 primary papillary adenocarcinomas reported that 20% were well differentiated, 68% were moderately differentiated, and 6% were poorly differentiated.
Metastatic tumors of the papilla are extremely unusual and may arise from a variety of primary lesions, including lymphomas, malignant melanomas, adenocarcinomas of the breast, and renal cell carcinomas.
Papillary adenomas, like adenocarcinomas, are histologically very similar to their colonic counterparts. One notable difference is that papillary adenomas are more likely to contain villous architecture. As in the colon, progressive dysplastic changes may occur in benign adenomas. Other pathologic categories of benign papillary tumors are important, primarily because they may be mistaken for malignancies and resected in a major operation.
In one series of 118 papillary tumors in patients undergoing surgical resection, 4 patients had tumors that were nonneoplastic and classified as papillitis or pancreatitis. Benign lesions, such as lipomas and leiomyomas, rarely are observed in the papilla.
Histopathologic studies of malignant tumors of the main duodenal papilla have demonstrated that these lesions are almost exclusively adenocarcinomas. Rarely, squamous and mucinous elements may be found within these malignant tumors. In addition, these tumors may be characterized as having papillary (villous) or tubular architecture, as well as a wide range of differentiation, from highly differentiated to well differentiated.
The incidence of stones in the common duct is increased in the setting of papillary cancers or adenomas. Whether this phenomenon is an indication that gallstones predispose people to papillary neoplasia or whether stones form in the bile duct as a result of stenosis is unknown. Various dietary carcinogens also have been postulated as possibly contributing to the formation of these tumors. As with other hypotheses concerning papillary tumor carcinogens, data are insufficient to either refute or support the propositions.
Investigators reported that disruption of the pRb-p16INK4 pathway may play a role in ampullary carcinogenesis, after examining the expression status of this protein and pRb immunohistochemically in 36 ampullary cancers. Methylthioadenosine phosphorylase (MTAP) plays an important role in the salvage pathway of adenosine, and inactivation of the MTAP gene is often noted in pancreatic cancer.
Hustinx et al reported that MTAP expression is lost in up to 30% of infiltrating pancreatic cancers and in a lower percentage of periampullary cancers, due to homozygous deletions of the MTAP and p16INK4A/CDKN2A genes.  The same investigators also reported that aberrant expression of certain proteins, such as 14-3-3 sigma and ERM proteins, is associated with a poorer prognosis. 
Although the etiology of papillary tumors is unknown, various risk factors have been identified; of these identified risk factors, cigarette smoking may account for 25-30% of cases. Other risk factors include diet (ie, high fat, high protein, low fruit, low vegetable intake), alcohol, coffee, occupation, the presence of other diseases (eg, chronic pancreatitis, diabetes mellitus, pernicious anemia), and prior gastric surgery.
From among these risk factors, aside from smoking, only chronic pancreatitis and adult onset diabetes mellitus of less than 2 years duration seem to be associated with an increased risk of pancreatic cancer. For example, chronic pancreatitis is associated with a 5- to 15-fold risk, whereas hereditary pancreatitis is associated with a 50- to 70-fold risk and a cumulative lifetime risk of 40% until age 75 years.
Pancreatic cancer may also occur in 3 other settings where there is an inherited predisposition. First, up to 10% of patients appear to have an inherited component in the absence of familial pancreatic cancer or other cancer syndrome. Second, there is an increased incidence of pancreatic cancer in individuals from families with familial pancreatic cancer in which the disease appears to be transmitted in an autosomal dominant manner with impaired penetrance. Studies have reported that 18% of these families may have BRCA2 mutations, in Jewish and non-Jewish populations.
Finally, an increased risk of pancreatic cancer may occur as part of another cancer syndrome, such as familial adenomatous polyposis (FAP), hereditary nonpolyposis colorectal cancer (HNPCC), Peutz-Jeghers syndrome, familial breast-ovarian syndromes, and familial atypical multiple mole melanoma. 
An adenoma-carcinoma sequence is postulated in individuals from high-risk groups, such as those with FAP. One prospective study of 18 patients with FAP or Gardner syndrome found 50% had adenomatous changes of the papillary mucosa; 2 of the 9 patients with adenomatous changes had normal-appearing papilla.
Causes of secondary papillary tumors include the following:
Renal cell carcinoma
Tumors of the papilla are unusual entities; in the United States, they make up less than 5% of all new digestive tract neoplasms. However, cancer of the pancreas is significantly more common, being responsible for approximately 12% of digestive organ cases. 
The prevalence of papillary adenomas in autopsy series was 0.04-0.12%, while a review of more than 1 million surgical pathology specimens discovered only 6 villous adenomas. Another report showed that only 12 of more than 400 pancreaticoduodenal tumors were papillary adenomas. The epidemiology of papillary adenomas is not significantly different from that of papillary adenocarcinomas.
A slight male predominance exists in the incidence of papillary tumors, although this may be beginning to change. Papillary tumors are uncommon before age 45 years; 80% of papillary tumors occur between the ages of 60-80 years.
The prognosis of adenocarcinoma of the papilla is the most favorable of all the periampullary tumors. The prognosis for patients with malignant tumors of the papilla depends on the extent of local tumor invasion and the presence or absence of vascular invasion, lymph node involvement, and distant metastases. [16, 17, 18, 19, 20, 21, 22, 23, 24]
In addition, Yokoyama et al reported, after evaluating 59 patients with ampullary carcinoma, that jaundice at presentation predicts an advanced stage ampullary carcinoma and portends a poor prognosis. 
For patients with node negative ampullary cancer, Brown et al reported pancreaticoduodenectomy was curative in 80% of patients and that once 3-year survival is reached, long-term survival can be expected. 
Patients with ampullary tumors that predominantly produce sialomucins have a better prognosis than do patients with tumors that secrete sulfated mucins, while patients with ampullary carcinomas that express carbohydrate antigen 19-9 or carcinoembryogenic antigen have a poorer prognosis than do those with ampullary carcinomas that do not express these antigens. 
Approximately 50% of ampullary lesions are deemed resectable. In 12 series reported since 1982 for patients with ampullary carcinoma who underwent resection, the overall mean 5-year survival rate was 40%, with a range of 16-61%. As the operative mortality rate continues to decrease and the selection of patients improves, the 5-year survival rate should improve accordingly.
Based on a review of available follow-up data, Schulten et al calculated a 5-year survival rate of 21% for patients with villous adenoma of the duodenum harboring malignant change. Celik et al concluded that radical excision of malignant, ampullary, villous adenomas in patients at good risk provides long-term survival in one third of these patients.
The 5-year survival rate for patients with carcinoid tumors without apparent metastasis who undergo resection was 68% and 95% in 2 different studies. Of note, the 5-year survival rates in patients with regional lymph node or distant metastasis are 83% and 38%, respectively.
Jaundice is observed in 80% of cases of papillary cancers. Unlike the jaundice observed with cancer of the pancreas, jaundice produced by papillary cancers may fluctuate, especially early in the course of the obstructive process.
The fact that jaundice tends to draw attention to ampullary tumors at a relatively early stage may account for the better cure rates seen with these tumors than for tumors further away from the papilla. Ampullary tumors may also bleed intermittently, resulting in occasional silver stools (due to the combination of melena with acholic stools). 
Other symptoms occurring with papillary tumors include the following:
Abdominal pain - Present in 37-60% of cases
Severe and rapid weight loss - Often associated with unresectability
Nausea and vomiting
Pancreatitis - Observed in 18% of patients
Migratory thrombophlebitis (Trousseau sign) - Rarely the first symptom of the disease
Virchow node is the presence of an enlarged supraclavicular lymph node and is a sign of unresectability. Courvoisier sign states that in the presence of jaundice, if a gallbladder is palpable, one must exclude biliary obstruction from malignancy.
Benign tumors present less often with jaundice, reflecting the smaller tumor size and absence of tissue invasion, although abdominal pain is common. One report of 25 patients with papillary adenomas found abdominal pain in 76% of cases, jaundice in 36%, and pancreatitis in 16%. The increased incidence of common duct calculi observed in benign and malignant tumors of the papilla may account for some of the abdominal pain and jaundice observed with these lesions.
In the series of Ryan et al, 16 of 19 patients with villous tumors of the duodenum had involvement of the main duodenal papilla. Villous adenomas of the ampulla range in size from 4 mm to 7 cm in diameter. Small tumors of the main papilla can be asymptomatic, but symptoms have been found to be present in more than 75% of reported cases. Signs and symptoms include the following:
Nonspecific abdominal pain (75% of cases)
Jaundice (70% of cases)
Upper gastrointestinal (GI) bleeding (50% of cases)
Coexisting gallbladder or common bile duct (CBD) stones (13-20% of cases)
Although most carcinoid tumors are asymptomatic and are found incidentally, those arising at the ampulla tend to cause symptoms related to intermittent biliary obstruction, including the following:
Carcinoid syndrome has not been reported in association with ampullary carcinoid tumors. This observation probably reflects the smaller size of these tumors when detected at the ampulla, a key factor in predicting the incidence of metastases.
Secondary tumors are rare. Patients present with GI bleeding, obstruction, and perforation. The treatment of secondary tumors is palliative.
Endoscopic Biopsy and Endoscopic Sphincterotomy
The yield of endoscopic forceps biopsies has ranged from 50-90%. The use of a polypectomy snare to obtain ampullary biopsies increases the size of the specimen as compared with the use of biopsy forceps. The risk for bleeding also is increased with the use of a polypectomy snare as compared with the use of standard forceps.
The location and the timing of endoscopic biopsies are important parameters in the diagnosis of ampullary carcinoma. Communicate the exact biopsy site to the pathologist because the morphology of the epithelium changes proximal to the papilla. Dark columnar cells that can be present (normally in the region of the papilla) could be interpreted as malignant in a specimen taken higher in the bile duct.
In ulcerated tumors, biopsies should be taken at the nodular edge and not from the base, because the latter consists of necrotic exudate. When biopsies are needed in a situation in which a sphincterotomy has been performed, they should be taken during the index procedure. Sphincterotomies may create histologic changes that last at least 6 days and mimic dysplasia.
Endoscopic sphincterotomy may play an important role in the diagnosis of ampullary tumors. However, performing an endoscopic sphinctertomy is advisable only when it is considered essential for diagnostic purposes or to eliminate the possibility of a tumor simulating stone impaction.
Narrowing of the biliary orifice after previous endoscopic biliary sphincterotomy has been referred to as sphincterotomy stenosis. Sphincterotomy stenosis is recognized late and is an uncommon complication of this procedure.
In these potentially curable lesions, endoscopic ultrasonography (EUS) can have the greatest clinical impact. Endosonographically, ampullary lesions appear as hyperechoic or mixed echogenic structures that spread to the duodenal wall, CBD, pancreas, and/or adjacent organs and vessels. However, EUS cannot distinguish a benign papillary adenoma from a papillary carcinoma unless infiltrative growth exists. 
A study evaluating preoperative staging of ampullary tumors found the overall accuracy of tumor staging to be 74% (26 out of 35) for all tumors. The accuracy for T1, T2, and T3 tumors was 67% (6 out of 9), 71% (10 out of 14), and 83% (10 out of 12), respectively. In the same study, EUS had an accuracy of 86% in helping diagnose pancreatic invasion. Because EUS is accurate for evaluating resectability of ampullary cancer, it can prevent unnecessary surgery and, thus, is critical in the management of ampullary cancers.
A retrospective study of EUS with fine needle aspiration biopsy (EUS-FNAB) of 35 patients with suspected primary ampullary tumors reported a sensitivity, specificity, positive predictive value, and negative predictive value of 82.4%, 100%, 100%, and 76.9%, respectively. The authors reported that adenocarcinomas were present in 13 patients (30%) and concluded that EUS-FNAB was an accurate test for assessing suspected primary ampullary tumors. 
Tumor-Specific Diagnostic Considerations
A study by Leese et al compared the characteristics of neoplasms with pseudotumors, the investigators reported that 49 patients of 2000 who underwent ERCP had the appearance of a periampullary carcinoma on duodenoscopy. Biopsies were taken before and after duodenoscopy, and ERCP was attempted by the insertion of a catheter through the tumor mass. 
Thirty-eight patients had neoplasms and 11 had inflammatory, nonneoplastic lesions that were termed pseudotumors. No statistical differences between these 2 groups were present with regard to any of the clinical features recorded, except that a higher proportion of patients with pseudotumors had epigastric pain. None of the patients with pseudotumors had a palpable gallbladder, whereas this finding was present in 16 patients with ampullary tumors.
Comparison of hematologic and biochemical tests revealed significantly lower values for hemoglobin, total protein, albumin, and alanine transaminase in the group with tumors. Comparison of ERCP results disclosed that even though the diameter of the pancreatic duct tended to be larger in the group with tumors, no difference existed with respect to the CBD diameter, and the cholangiograms appeared similar in many cases. Gallstones were present significantly more often in the pseudotumor group.
Two of the 11 patients with pseudotumors were subjected to surgical excision because of the histologic features on endoscopic biopsies that raised the suspicion of carcinoma. Thus, although certain clinical, laboratory, and endoscopic biopsies raised the suspicion of carcinoma, significant overlap with the findings in benign disease and malignant disease existed.
When endoscopic biopsies are equivocal for carcinoma and the suspicion of a pseudotumor exists, repeating duodenoscopy with biopsies in 2-4 weeks is indicated. A significant proportion of pseudotumors, especially those relating to gallstone disease, have resolved in this time period, thus preventing unnecessary surgery.
In the presence of benign tumors, laboratory studies are consistent with cholestatic jaundice (increased alkaline phosphatase, gamma-GT, and direct bilirubin). Imaging studies occasionally reveal a filling defect in the duodenum.
Villous adenomas of the ampulla have a high incidence of coexisting malignancy. In a surgical series, the incidence of malignant change (carcinoma in situ and invasive carcinoma) ranges from 26-63%. Any assessment of the prevalence of malignancy within ampullary villous adenomas must be based on histopathologic analysis of the entire specimen, because endoscopic biopsies have an unacceptably high false-negative rate. In one series, endoscopic examination found the tumor in 9 of 10 patients, but endoscopic biopsies missed the rate of malignant change in 5 of 9 patients, or 56%.
No distinctive endoscopic features exist for neuroendocrine tumors, but CT scanning, magnetic resonance imaging (MRI), and EUS can be employed in the diagnosis. [32, 33] The tumor may spread locally to the duodenal wall, pancreas, or bile duct, or it may spread by metastases to local and regional lymph nodes. Lymph node metastases are observed in 25-50% of cases.
Given the high false-negative rate for endoscopic biopsy in detecting foci of malignancy, local surgical excision and pancreaticoduodenectomy have been the mainstays of treatment of villous adenomas of the main duodenal papilla. The selection of a surgical procedure primarily depends on the results of endoscopy and intraoperative frozen-section biopsies. For tumors without evidence of malignancy, submucosal excision is the procedure of choice, and this is combined with sphincterotomy or sphincteroplasty if the tumor is growing within the ampullary orifice. More than 85% of patients with benign villous adenomas treated with local excision do well clinically and remain free of recurrence.
If intraoperative frozen sections or examination of the local excision specimen reveals carcinoma in situ, most authorities recommend radical pancreaticoduodenectomy, although others, especially in older reports, have recommended a conservative approach with local excision alone. For lesions that contain invasive carcinoma, the operation of choice is pancreaticoduodenectomy.
Local excision and pancreaticoduodenectomy are the 2 operations most commonly employed in the management of neuroendocrine tumors. Although available data are limited, pancreaticoduodenectomy may be assumed to be associated with a higher perioperative morbidity and mortality.
In patients who are carefully selected, a place exists for reoperation of recurrent periampullary adenocarcinomas at a referral center for pancreaticobiliary surgery.
The role of endoscopy in the management of tumors of the main duodenal papilla continues to evolve. Preoperative endoscopic drainage improves liver function and nutrition, with additional opportunity to obtain biopsies if the initial diagnosis is equivocal. Endoscopic therapy is associated with a much lower mortality rate than are surgical bypass procedures. [34, 35, 36, 37]
Surgical excision remains the mainstay of treatment of adenocarcinomas of the main duodenal papilla in patients who are at good risk without metastatic disease. Bettschart justified an aggressive surgical approach to ampullary lesions based on the overall low proportion of benign lesions, absence of postoperative mortality, and improved long-term survival in 561 patients with periampullary tumors. [38, 39, 40, 41, 42, 43]
If the tumor is deemed resectable, the operation of choice is a pancreaticoduodenectomy. Although the Whipple version has been the most popular procedure, the development of pylorus-preserving pancreaticoduodenectomy has provided a reportedly easier and less time-consuming operation with less blood loss, a shorter hospital stay, and better weight gain during follow-up care. Also, no differences in the recurrence rate and patient survival exist between pylorus-preserving pancreaticoduodenectomy and the standard Whipple procedure. The operative mortality rate for the Whipple procedure has declined from 26% to less than 5%, although the associated mortality rate remains in the range of 30-60%.
Beger et al reported 5-year survival as between 45-65% with pancreaticoduodenectomy. Factors associated with a long-term survival included absence of lymph node involvement and absence of infiltration into the pancreas. 
Jarufe et al, reporting their experience with 251 patients with pancreatic carcinoma, found that patient survival was influenced by poor tumor differentiation, surgical margin, lymph node metastases, and age, with an overall mortality of 4.8% and a morbidity of 29.9%. Tumors in the study included 133 pancreatic adenocarcinomas, 88 ampullary tumors, and 30 distal CBD tumors. 
Kim et al reported that the predictors of failure after pancreaticoduodenectomy were the staging of disease and nodal status. 
Regional pancreatectomy, an operation developed by Fortner, involves the en bloc removal of a tumor in or immediately adjacent to the pancreas, with an adequate soft-tissue margin and regional lymphatic drainage of the tumor. This operation may allow for a significantly higher resectability (compared with the Whipple procedure) of ampullary carcinomas with vascular involvement because it involves resection of the pancreatic segment of the portal vein; resection of involved arterial structures, if necessary; and a total or subtotal pancreatectomy.
An argument in favor of local surgical excision for small ampullary adenocarcinomas can be made only for patients who are not fit to undergo a pancreaticoduodenectomy. Although formal guidelines for follow-up with patients who have undergone local excision of the ampullary tumors do not exist, duodenoscopic surveillance is logical because the recurrence rate is not insignificant. Semiannual examination for 1-2 years followed by annual examination is a suggested surveillance schedule.
However, ampullectomy for early ampullary cancer is not recommended, because of the high risk of recurrence. Ampullectomy should be recommended only for patients with small lesions (pTis or pT1 1cm or less with high operative risk for pancreaticoduodenectomy). 
A palliative, proximal, biliary-enteric and gastrojejunal bypass procedure often can be performed in patients who are unfit for radical procedures and who have large bulky tumors that obstruct a significant portion of the duodenal lumen and the bile duct. The operative mortality rate of patients with ampullary tumors undergoing palliative surgical bypass ranges from 17-60%, with a mean of 32%. For patients with inoperable disease without significant duodenal obstruction, consider endoscopic palliation in lieu of surgical bypass because the morbidity and mortality are lower.
Endoscopic sphincterotomy is an effective means of palliating patients with ampullary adenocarcinomas and is associated with a shorter hospital stay, fewer complications, and a lower 30-day mortality rate than are surgical bypasses.
Sphincterotomy-associated biliary strictures are not amenable to repeat sphincterotomy. However, studies have suggested that they may be managed successfully by serial placement of stents of incrementally increasing diameter. Hydrostatic balloon dilatation could have been considered as an alternative therapy; however, these strictures, though short, were high grade and recalcitrant to treatment, necessitating prolonged periods of stent placement (median of 12.5mo) to achieve resolution. Serial placement of stents of incrementally increasing diameter offers sustained radial expansion of the stricture over time, in contrast to the short duration of the effect of the balloon. Consider surgery in cases resistant to endoscopic therapy.
Shemesh et al successfully used endoscopic diathermic fulguration against recurrent papillary adenomas. The investigators reported on 5 patients with adenomas of the papilla who presented with obstructive jaundice and underwent successful drainage via endoscopic sphincterotomy. Four of these patients underwent local surgical excision, with subsequent adenoma recurrence in all 4 within 6-18 months of the operation.
The recurrent adenomas were eradicated by endoscopic diathermic fulguration using the tip of a polypectomy snare or a hot biopsy forceps; no recurrences were noted during a follow-up period of 12-24 months. One patient with a sessile adenoma of the papilla refused surgery and, therefore, underwent incomplete piecemeal resection of the adenoma by snare polypectomy; an adenocarcinoma at the head of the pancreas developed 40 months later.
Binmoeller et al described 25 patients with adenomatous tumors of the main papilla who underwent radical endoscopic resection by snare papillectomy. The 3 criteria for inclusion in this study were tumor size less than 4 cm, benign endoscopic appearance, and benign histologic results on at least 6 forceps biopsies. Twenty-three patients had de novo tumors, and 2 had recurrent adenomas after local surgical excision. 
Endoscopic excision was performed with a pure-cutting current to the level of the muscularis propria. Residual tissue was fulgurated using a monopolar current. ERCP was performed after the papillectomies; then biliary or pancreatic duct papillotomies, or both, were performed, as well as endoprosthesis placement in some patients (as satisfactory drainage was needed). Surveillance duodenoscopy was performed at 1, 6, and 12 months after the papillectomies and yearly thereafter. Immediate complications included postpapillectomy bleeding in 2 patients, necessitating local injection of epinephrine (1:20,000), and self-limited pancreatitis in 3 patients. No procedure-related deaths occurred.
Two patients had evidence of intraductal tumor extension by ERCP and were referred for surgery. The remaining 23 patients were observed for a median of 37 months. Six patients had benign recurrences (26%), usually within the first year of follow-up. Of the 6 patients with recurrences, 1 patient had tumor extension into the distal CBD and later succumbed to complications of a Whipple procedure.
The remaining 5 patients with recurrence of adenomas were treated endoscopically with snare resection (1 patient), diathermic fulguration (2 patients), or combined treatment (2 patients). Three of these 5 patients were free of disease at 13, 52, and 53 months. One patient was lost to follow-up, and 1 patient underwent a pancreaticoduodenectomy after repeated fulgurations failed to completely ablate adenomatous tissue.
Silvis correctly urged caution and careful consideration of several points in the above study before embarking on endoscopic papillectomy of benign ampullary tumors. First, the patients in the study by Binmoeller and colleagues were unusual in that no patient had cancer within the adenoma, whereas the incidence in other reports has ranged from 26-63%. 
Second, papillectomy is associated with a significant recurrence rate and, therefore, should be considered only for patients who are likely to be compliant with an extended schedule of follow-up endoscopic surveillance procedures. Third, a papillectomy only should be performed by an endoscopist who is an expert in ERCP. Lastly, this procedure should be restricted to either patients who refuse surgery or patients who are at high risk for extensive surgical procedures.
Although the issue of stent placement for palliative management of ampullary adenocarcinomas remains unsettled, patients who are discharged from the hospital with an endoprosthesis should be instructed to seek medical attention immediately if they develop any signs or symptoms of biliary obstruction or cholangitis. 
Consider permanent placement of an expandable metal stent in patients with inoperable ampullary adenocarcinomas or in patients who are not fit for surgery. For patients who have developed stent occlusion and recurrent jaundice as a result of tumor growth, insertion of a plastic prosthesis into the occluded metal stent or thermal ablation of the tumor has been employed. Metal stents have 2 major disadvantages in that they cannot be removed and they are relatively expensive. Stents should not be placed in patients who may be surgical candidates.
What would you like to print?
- Tumor Characteristics
- Patient History
- Endoscopic Biopsy and Endoscopic Sphincterotomy
- Endoscopic Ultrasonography
- Tumor-Specific Diagnostic Considerations
- Surgical Overview
- Local Excision
- Biliary Bypass
- Endoscopic Sphincterotomy
- Diathermic Fulguration
- Snare Papillectomy
- Stent Placement
- Show All