Endoscopic submucosal dissection for duodenal tumors including papilla: Is it feasible?

In recent years, the incidence of superficial nonampullary duodenal epithelial tumors (SNADETs) detected using upper gastrointestinal endoscopy has increased in asymptomatic individuals. Most SNADETs are adenomas or intramucosal adenocarcinomas for which minimally invasive endoscopic treatment is often the preferred therapeutic approach, underscoring its growing significance. However, SNADETs are less prevalent than other gastrointestinal neoplasms such as those of the esophagus, stomach, and colorectum, and endoscopic diagnostic and therapeutic modalities for these tumors remain in the developmental stage.

Endoscopic submucosal dissection (ESD) is recommended when en-bloc resection is difficult to achieve using reliable endoscopic mucosal resection (EMR) or underwater EMR (UEMR). This technique is frequently selected for SNADETs of >2 cm in size, those with scars, or those with anatomically curved locations because the en-bloc resection and R0 resection rates of ESD are higher than those of cold snare polypectomy, EMR, and UEMR.¹ Although duodenal ESD has been reported to have an extremely high incidence of adverse events in the past,² improvements in various resection techniques and devices, such as the water pressure method, pocket-creation method, and ESD using scissors-type forceps, have made it possible to reduce intraoperative adverse events.^3-5 With regard to delayed adverse events, reliable wound closure up to 3 days after endoscopic resection is important to prevent adverse events because wound closure after resection reduces the risk of delayed adverse events,¹ and delayed perforation occurs within 3 days after endoscopic treatments.⁶

In this issue of Digestive Endoscopy, Yahagi et al.⁷ conducted a retrospective cohort study of patients who underwent ESD for duodenal tumors including the papilla (ESDIP). This technique was primarily introduced for the treatment of lesions of >20 mm in size and cases with laterally spreading morphology, in which it is difficult to achieve en-bloc resection by endoscopic papillectomy (EP). The en-bloc resection rate of ESDIP is extremely high and the R0 resection rate is relatively low (96% and 46%, respectively). Furthermore, the rate of perioperative adverse events was high (intraoperative perforation, 15%; postprocedural bleeding, 19%; and post-ESDIP pancreatitis, 25%). The findings indicated that while ESDIP had a high likelihood of achieving en-bloc resection, there is a risk of perioperative adverse events.

ESDIP presents a significant technical challenge, and is associated with a high risk of complications. Although Yahagi et al. performed the entire ESDIP procedure using water pressure method,³ which decreases intraprocedural perforation during duodenal ESD, the high rate of intraoperative perforation reflects the difficulty of performing ESDIP. Therefore, substantial experience and expertise in duodenal ESD are required for ESDIP. Moreover, insertion of an endoscopic nasobiliary pancreatic drainage (ENBPD) tube is necessary to reduce the risk of delayed perforation, because its rate after ESDIP is lower than that after SNADET. Conversely, even if ENBPD insertion was successful in all cases of endoscopic retrograde cholangiopancreatography (ERCP), the rate of post-ERCP pancreatitis (PEP) was high in this study, but similar to the reported incidence after duodenal ESD with ENBPD (16.0%).⁸ These results suggest that ENBPD insertion has a high risk of PEP, regardless of dissecting the sphincter muscles of the papilla. In any case of ESDIP, support from an experienced biliopancreatic endoscopy team is essential. Additionally, collaboration between hepatobiliary and pancreatic surgical teams capable of performing pancreaticoduodenectomy (PD) in emergency situations is imperative. Fortunately, cases that requires ESDIP are rare; therefore, it is unnecessary for every institution to perform ESDIP. It is advisable to refer patients to advanced institutions as needed, after comprehending the characteristics and indications of the procedure.

EP for ampullary adenomas of up to 20–30 mm without bile duct or pancreatic duct invasion is widely accepted in Eastern and Western countries. PD remains the standard treatment option for ampullary adenocarcinomas. EP may be acceptable for Tis adenocarcinomas but is not recommended for T1 cases, which have a 10% incidence of lymph node metastasis. Most ampullary tumors do not exceed 20–30 mm in size, and EP can be performed in many high-volume centers in Japan. However, the recurrence rate of EP is likely to be up to 23%.⁹ Moreover, local recurrence due to positive margins in both the horizontal and vertical directions is expected to increase when EP is performed for ampullary tumors exceeding 20–30 mm in size. Therefore, PD is usually recommended for widespread lesions outside of the papilla. Regardless of the large tumor size (mean 39 ± 21 mm), ESDIP achieved a higher en-bloc resection rate than EP. The overall survival rate was excellent (96%), although there was a 15% local recurrence rate with additional endoscopic and surgical treatment of adenoma and adenocarcinoma, respectively. Therefore, ESDIP may be an alternative option to PD for ampullary lesions exceeding 20–30 mm in size, if its safety and curative potential are ensured.

This study showed that ESDIP is associated with the risk of local recurrence due to a positive pathological margin. It is important to understand that the majority of positive vertical margins are present in the papillary region, and that any lesions that require ESDIP have a risk of intrapapillary extension, which may explain the low R0 resection rate. The performance of endoscopic ultrasonography (EUS) in the preoperative diagnosis of ampullary tumors was evaluated in a meta-analysis.¹⁰ The pooled sensitivity and specificity of EUS were 77% (95% confidence interval [CI] 69–83%) and 78% (95% CI 72–84%), respectively. The diagnosis of infiltration into the sphincter of Oddi by EUS remains difficult.

In conclusion, this study demonstrates that ESDIP is technically feasible and that complications remain prevalent only in a specialized high-volume center. ESDIP may be an alternative to PD for ampullary lesions exceeding 20–30 mm in size. Therefore, further technical and device improvements are essential to ensure the safety of ESDIP. Moreover, the analysis of the differential indications for ESDIP and EP has been insufficient. Consequently, future research should focus on accumulating ESDIP cases as well as data on their long-term outcomes.

Authors declare no conflict of interest for this article.

None.