To identify predictors of efficacy and safety following endoscopic papillectomy (EP), including therapeutic success, recurrence and adverse events. Secondly, we aimed to evaluate predictors of histopathological agreement.

A multicenter, retrospective cohort study was conducted including adult patients from thirteen medical institutions who underwent EP for ampullary adenomas between January 2019 and December 2023. The main clinical outcomes were efficacy (endoscopic success and recurrence) and safety (reported adverse events, severity and management). Pre- and post-procedural histology was documented. Patient center volume was categorized as low, medium and high. 

This cohort included 187 patients, 59.9% male, mean age 62.64 (± 13.41) years. Most of the ampullary lesions were not associated with genetically predisposed diseases (77.0%), incidentally detected on asymptomatic patients (76.5%), sized between 10mm and 20mm (63.1%), and with low-grade dysplasia (88.8%). En bloc resection was performed in 73.8% of the patients. Prophylactic measures to prevent acute pancreatitis were implemented in 95.2% of the procedures. Endoscopic success rate was 96.3%, with R0 resection rate of 74.3%. The mean follow-up was 32.8 (± 18.7) months. The median adenoma-free survival was 24.0 months (95% CI 9.3–38.7).  Recurrence rate was 18.7%, most frequently detected at the first endoscopic surveillance (71.4%) and managed mostly endoscopically (85.7%). Re-recurrence rate was low (4.3%, n=8). Older age (OR=0.95, 95% CI 0.91–0.99), intraductal extension (OR=17.89, 95% CI 1.51–212.43), intraprocedural bleeding (OR=3.55, 95% CI 1.34–9.39), and submucosal injection before resection (OR=2.60, 95% CI 1.06–6.40) were significantly associated with higher recurrence rates. The most common adverse events were post-procedural pancreatitis (20.3%), mostly mild severity (78.9%), and bleeding (intraprocedural 18.2%; delayed 12.8%). No procedural-related deaths were observed in our cohort. None of the patient, procedure or lesion characteristics were related to post-procedural pancreatitis, cholangitis, papillary stenosis or perforation. Delayed bleeding was more common when submucosal injection was performed (OR=2.96, p=0.029) and pancreatic stent was placed (OR = 5.05, p=0.036). The histopathological agreement between pre-procedural biopsy and post-procedural histology was observed in 75.4% of cases, with histopathological underestimation in 15% and overestimation in 9.6% of cases. Prior sphincterotomy correlated with reduced histopathological agreement (χ² = 21.81; p < 0.001), whereas pre-procedural evaluation with side-viewing endoscopy was associated to a higher likelihood of histopathological agreement (χ² = 11.10, p = 0.004). High-volume centers showed significantly higher rates of histopathological agreement (χ² = 15.59, p<0.001). It is important to note that, in these centers, prior assessment with side-viewing endoscopy and endoscopic ultrasound (EUS), respectively [(97.3% (n=72) and 71.6% (n=53)] was more often performed. Significantly higher recurrence rates (31.0%) were observed in low-volume centers compared with high-volume centers (12.2%) (p = 0.046). Delayed bleeding occurred more frequently in low-volume centers (p = 0.032).

In our cohort, EP achieved high rates of therapeutic success with a favorable profile of adverse events, mostly amenable to conservative or endoscopic management. Recurrence was frequent and was associated to older age, intraductal extension, intraprocedural bleeding and submucosal injection before resection. High-volume centers showed significantly higher rates of histological agreement probably due to the more frequent use of prior evaluation with side-view endoscopy and EUS.  Low-volume centers showed higher rates of recurrence rates and delayed bleeding. These results emphasize the importance of detailed pre-procedural staging, specialized training and the potential benefit of centralizing complex procedures to high-volume centers.