Outcomes of endoscopic papillectomy should be evaluated not only based on short-term results but also long-term prognosis

Abstract

Endoscopic papillectomy (EP) for ampullary tumors has become a widely performed treatment as the number of accumulated cases has increased across multiple institutions.¹ However, due to the anatomical characteristics where both the bile duct and pancreatic duct open, pathological assessment of resected specimens remains challenging, making the evaluation of treatment outcomes difficult. In other gastrointestinal tumors, margin assessment can be conducted based on horizontal and vertical directions alone, whereas in ampullary tumors, margin evaluation in the direction of the bile duct and pancreatic duct is also necessary. Additionally, the generally small size of resected specimens further limits the pathological margin assessment following EP. This contributes to the broad range of reported R0 rates, varying significantly from 47% to 93% in previous reports.² Guidelines for EP have been established in the United States, Europe, and Japan. Of the three guidelines, those of the European Society of Gastrointestinal Endoscopy³ and the Japanese² state that ampullary adenomas without intraductal extension are ideal indications for EP.

Binda et al.⁴ and The Interventional Endoscopy and Ultrasound group conducted a retrospective study on 430 EP cases to establish a scoring system predicting cases at high risk for incomplete resection (IR) based on preoperative factors. IR was defined as cases where lateral or endoampullary margins were affected by residual tumor post-EP, meaning they were not diagnosed as R0 pathologically. Despite including many cases considered high risk for IR, such as 60 cases (14.0%) of T1 or higher adenocarcinoma, 68 cases (15.8%) of intraductal extension (IDE) smaller than 20 mm, and 83 cases (19.3%) of laterally spreading tumors (LST) larger than 10 mm, the reported IR rate was relatively favorable at 23.6% (99 cases) compared to previous reports. This study introduced the PANETH score, with a common bile duct dilatation (CBD) diameter of 8 mm or more without cholecystectomy or 10 mm or more after cholecystectomy as 1 point, and the presence of IDE and LST as 2 points each, for a total of 3 points or more as significant risk factors for IR. Although adenocarcinoma itself was presumed to be a significant risk factor for IR, it was not included in the multivariate analysis to avoid confounding with the CBD factor. In our experience as well, cases with CBD or elevated biliary enzymes often result in final pathological diagnoses of adenocarcinoma, supporting the validity of this assessment.

In this study, 95 patients required additional surgery or were in a condition necessitating surgery. Additionally, 132 patients experienced recurrence. While there may be some overlap between these groups, the total number exceeds the reported 99 IR cases. This suggests that some cases deemed R0 still required additional surgery or recurred. However, the report does not analyze risk factors for additional surgery or recurrence.

In our single-center retrospective study⁵ of 212 cases with at least 6 months of follow-up after EP or additional treatment (including 53 cases of T1a(M) or lower adenocarcinoma), multivariate analysis identified IDE (P = 0.005) as the only significant risk factor for requiring additional surgery (additional surgery was required in 15.0% and 2.6% of cases with and without IDE, respectively). In contrast, the presence of adenocarcinoma (P = 0.288; additional surgery rate, 9.4%) and positive/uncertain pathological margins (P = 0.995; additional surgery rate, 5.7%) were not significant. Additionally, significant risk factors for recurrence included female sex, familial adenomatous polyposis, and IDE (P-values 0.004, <0.001, <0.001, respectively; recurrence rate, 18.1%, 38.5%, 35.0%, respectively). However, positive/uncertain margins (P = 0.58; recurrence rate, 15.1%) were not significant. In another retrospective study⁶ limited to 67 cases of T1a(M) or lower well-differentiated adenocarcinoma treated with EP, different findings emerged. Positive adenoma or adenocarcinoma margins (P = 0.010 in univariate analysis) were the only significant risk factors for additional surgery and IDE (P = 0.098) was not significant (additional surgery rate, 27% and 30%, respectively). These findings suggest that while pathological margin assessment should not be disregarded, R0 status alone does not guarantee the absence of recurrence.

Regarding LST, recent reports have evaluated the outcomes of endoscopic treatment, including endoscopic submucosal dissection (ESD) techniques, for duodenal tumors involving the papilla, performed in high-volume centers. LST is primarily associated with horizontal margin involvement leading to IR, and ESD may improve negative margin rates, ultimately enhancing prognosis. Yahagi et al.⁷ reported on 54 cases of duodenal tumors, including those involving the papilla, with a mean tumor size of 39 mm. Their study found a high horizontal negative margin rate of 92%. However, the papillary margin (likely referring to the bile duct and pancreatic duct margins) had a much lower negative rate of 54%, reducing the overall R0 rate and contributing to a 12-month cumulative recurrence rate of 12%.

For IDE cases, a multicenter prospective study by Camus et al.⁸ evaluated intraductal radiofrequency ablation (ID-RFA) for IDE shorter than 20 mm in 20 patients, reporting residual neoplasia rates of 15% at 6 months and 30% at 12 months. Based on these findings, the European Society of Gastrointestinal Endoscopy guidelines³ suggest that EP can be performed for IDE cases shorter than 20 mm using complementary techniques such as ID-RFA. Moving forward, instead of avoiding EP in cases predicted to have IR based on pathological margin concerns, efforts should be made to apply novel techniques and assess their safety, additional surgery rates, recurrence rates, and long-term outcomes, even in cases with a PANETH score of 3 or more.

Conversely, for adenoma cases, prognosis may not necessarily worsen even without treatment. Given that EP still carries a risk of severe complications, patient age and comorbidities must be considered when determining indications. Our single-center retrospective study⁹ analyzed 236 EP cases in patients aged 37–85 years, excluding familial adenomatous polyposis cases. The study identified an age-adjusted Charlson Comorbidity Index score of 5 or higher as the only significant factor worsening prognosis. In contrast, IDE (P = 0.057) and the presence of T1a(M) or lower well-differentiated adenocarcinoma (P = 0.923) were not significant prognostic factors. These results suggest that in elderly patients with severe comorbidities, careful indication determination and thorough informed consent are necessary.

In conclusion, while pathological margin assessment in EP is crucial, it is not always definitive, and undue emphasis should be avoided. Although surgical ampullectomy is an alternative for IDE and LST cases unsuitable for normal EP, it is not a standardized procedure, has a higher invasiveness than EP, and experienced surgeons are limited. Instead, efforts should be made to eliminate residual lesions using endoscopic techniques such as additional endoscopic treatment, ID-RFA,¹⁰ and ESD. Furthermore, patient age and overall condition should be considered when determining indications to ensure safe EP that improves long-term prognosis. Future research should continue accumulating evidence on EP indications and techniques. Evaluations should not only focus on short-term outcomes such as R0 rates and complications but also consider long-term prognosis.

Authors declare no conflict of interest for this article.

None.