To evaluate and compare the results of endoscopic ampullectomy (EA) and surgical ampullectomy (SA) for benign tumors.

From 2005 to 2018, 66 patients were eligible for ampullectomy (adenoma, ultrasound grading≤T1, N0). The principal endpoint was the rate of complete resection: R0 resection without local recurrence and no recurrence after a single operative procedure and with final pathology confirming the operative indications. Secondary endpoints were: morbidity, mean length of stay (LOS), preoperative and final pathology results, complete (or incomplete) resection, mean duration of follow-up, local recurrence rate and late complications (biliary stricture).

Among the 41 patients eligible for EA, 36 ampullectomies were performed, while five were referred for SA because of intracanalar invasion that had not been diagnosed initially. The rate of complete treatment was 48% (13/27), the morbidity was 10%; the LOS was 3.3days; no adenomatous lesion was found on the definitive pathology in 9 patients (25%). The resection was R1 in 30% but there were no R2 resections. The median duration of follow-up was 20months, the rate of local recurrence was 22% and the rate of late biliary stricture was 2%. Among the 30 patients who were eligible for SA, (25 plus the five referred by endoscopists after intracanalar invasion was found), SA was eventually performed in 24: five patients were converted to pancreatoduodenectomy, one patient was found to have metastatic disease and had no resection. An extemporaneous frozen-section pathology examination was performed in 22 of the 24 SA patients and confirmed healthy margins on the bile ducts and pancreatic ducts. The rate of complete treatment was 71% (17/24), morbidity was 35%; LOS was 10days; seven patients (29%) were found to have invasive adenocarcinoma, and six of these (86%) underwent subsequent pancreatoduodenectomy. None of the resections were R1 or R2; median follow-up was 21months; there was no local recurrence; biliary stricture rate was 8%.

The short-term and long-term results of patients undergoing endoscopic or surgical ampullectomy are different. The lesions resected by EA were less advanced, with simpler postoperative course. The lesions for which SA was performed were more advanced, and had more morbid sequelae, but with a better rate of complete treatment and better long-term results.