Endoscopic Papillectomy: Introduction and How to Treat
Ampullary tumor is a comparatively rare tumor derived from the duodenal papilla. It is thought to be associated with the progression of adenoma to carcinoma and is recognized as a premalignant lesion. Recently, endoscopic papillectomy, snaring resection of the ampulla, has been accepted as a less invasive alternative to surgical treatment for cases of ampullary adenoma or adenocarcinoma in adenoma in patients for whom curative resection was possible. The rates of curative resection were reported to be 52–92%, with 0–33% of recurrence rates. However the indications, preprocedural diagnosis, and technique of endoscopic papillectomy are still not standardized. Experts agree that endoscopic ultrasonography is useful, for the diagnosis on the presence of invasion to the muscularis propria, intraductal extension of the lesion, and metastasis to regional lymph nodes. Pancreatitis and bleeding are the most common complications. Pancreatitis is considered to occur due to obstruction of pancreatic duct orifice and thermal damage to the pancreatic parenchyma. Pancreatic stent placement is recommended to avoid obstruction of pancreatic duct orifice. The endoscopic surveillance after endoscopic papillectomy is essential for detecting recurrence of ampullary adenoma.
KeywordsEndoscopic papillectomy Ampullary adenoma Endoscopic retrograde cholangiopancreatography Endoscopic ultrasonography
33.2 The Role of Endoscopic Papillectomy
Classically, surgical resection (local resection or pancreaticoduodenectomy) is the standard treatment for ampullary adenoma. Surgical resection has the advantage of a low recurrence rate, but it is too invasive for cases of localized ampullary adenoma. Currently, endoscopic papillectomy (EP) has been accepted as a less invasive alternative to surgical treatment for cases of ampullary adenoma in patients for whom curative resection was possible.
EP was first documented by Suzuki and Murakami in 1983 . It involves the resection of the mucosa and submucosa of the duodenal wall, in the area of the anatomical attachments of the ampulla of Vater, including the tissue around the bile duct and the pancreatic duct orifices . Curative resection of EP is reported to be achieved in 52–92% of cases [8, 9, 10, 11, 12, 13, 14, 15]. The complication rate is reported as being between 9.7 and 20%, with common complications including post-procedural pancreatitis (0–25%), bleeding (0–25%), perforation (0–8%), and cholangitis (2–3%) [8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]. At present, EP is considered a less invasive option to surgery, but as a high-risk endoscopic procedure. Therefore the procedure requires specialist expertise. Careful observation after the procedure is important to detect acute complications.
The indications for EP are still not standardized. From previous reports, EP is accepted for patients with ampullary adenomas smaller than 4–5 cm  without ductal extension. Intraductal involvement of the lesion is considered to signify noncurative lesions or those with a high risk of recurrence . Adenocarcinoma in adenoma within the mucosal layer is reported to be suitable for EP; there is a rare risk of lymph node metastasis , and the misdiagnosis of malignancy on endoscopic forceps biopsy is frequent [13, 22].
33.3 Preprocedural Evaluation for Endoscopic Papillectomy
Recent advancements in endoscopes, endoscopic retrograde cholangiopancreatography (ERCP), and endoscopic ultrasonography have contributed to the development of EP. In spite of all this, the accuracy of preprocedural diagnosis is reported only by experts.
The diagnosis of ampullary tumor is performed through endoscopic appearance, biopsy, EUS, and ERCP with intraductal ultrasonography (IDUS). Extracorporeal US, computed tomography (CT), and magnetic resonance cholangiopancreatography (MRCP) are useful for the detection of large lesions, metastasis, and indirect findings (biliary and pancreatic duct dilation). However, they are inappropriate for the evaluation of small lesions.
33.3.1 Endoscopic Appearance and Pathological Diagnosis
The typical endoscopic finding of ampullary adenoma is a villous tumor. The characteristic feature distinguishing between adenoma and adenocarcinoma is the presence of ulceration, which has been observed in patients with malignancy, but never in patients with benign disease. The fold convergence of the duodenum wall around the ampulla indicates tumor invasion into the duodenal wall. However, ampullary adenomas cannot always be distinguished from ampullary carcinomas according to endoscopic appearance alone. Observation of the ampullary tumor with narrowband imaging (NBI) is reported to be helpful for providing endoscopic images of microvessels and the surface structure of tumors  or to enhance tumor margins .
Biopsy is very important in differentiating adenoma from carcinoma or other tumors. Pancreatitis may occur following biopsy. The orifice must be carefully confirmed when taking a biopsy from a small lesion. However the accuracy of biopsy is reportedly not high, around 70% [11, 13, 22, 25]. It is thought that severe atypism is observed in the ampullo-pancreatobiliary common duct rather than in the ampulla-duodenum. Therefore, biopsy must be taken from the deep portion of the orifice. When adenocarcinoma is suspected, biopsy followed by endoscopic sphincterotomy  or endoscopic ultrasonography-guided fine needle aspiration is considered. However, it is also reported that the sensitivity of biopsy did not change after sphincterotomy . In these results, endoscopic papillectomy is sometimes performed as a major biopsy prior to surgery.
33.3.2 ERCP, IDUS, and EUS (Fig. 33.2)
EUS is essential for deciding whether or not endoscopic resection is indicated. The EUS provides information on the presence of invasion to the muscularis propria, intraductal extension of the lesion, and metastasis to regional lymph nodes. EUS is reported as being superior to CT, magnetic resonance imaging, or transabdominal ultrasonography as a diagnostic modality [28, 29]. In meta-analysis, the pooled sensitivity and specificity of EUS in the diagnosis of T1-stage tumors were 77 and 78% . Ridtitid et al. reported that, in a retrospective cohort study of patients with ampullary tumors, the overall accuracy of EUS for the assessment of tumor extent was comparable to ERCP. The authors concluded that ERCP and attempts at endoscopic resection of the ampullary tumor should be avoided in selected cases of local tumor invasion or intraductal extension detected with EUS .
ASGE guidelines recommend that ERCP with both biliary and pancreatic duct evaluation should be performed at the time of endoscopic resection to assess for evidence of extension into either ductal system, especially in cases where EUS is not performed . IDUS is inserted through the working channel of the jejunoscope and into the bile and pancreatic ducts after cholangiopancreatography. IDUS may be useful for imaging the detailed anatomy of the ampulla of Vater than EUS. Ito et al. reported that the combination of EUS and IDUS made better accuracy of preprocedural diagnosis . There are some opinions, however, that all the lesions do not require evaluation with EUS and IDUS before endoscopic therapy, because small lesions (<1–3 cm) without suspicious malignancy have low risk of muscular invasion or ductal extension.
33.4 Techniques of EP (Fig. 33.3)
Achieving en bloc resection without complication is fundamental in performing EP. Complete pathological evaluation is important for the evaluation of the resected margins or malignant foci with invasiveness, as previous pathological diagnosis is often incomplete. In a few case reports, balloon-catheter-assisted papillectomy was documented to facilitate en bloc resection [33, 34]. Piecemeal resection is performed for large lesions, which aims to decrease complications and recurrence. However, histopathological evaluation of the resected margin is then difficult. There are no data comparing safety or recurrence rates between en bloc and piecemeal resections.
33.4.1 Submucosal Lifting
Submucosal saline injection with or without indigo carmine prior to EP has been performed in some reports [15, 17]. Submucosal lifting may reliably indicate malignancy, may prevent the effect of electrosurgical current, and therefore may prevent post-procedural pancreatitis. It may be useful for cases with predominant lateral periampullary extension . However, the mucosal tissue at ampullary lesions does not lift because of tethering by the biliary and pancreatic ducts. Additionally, the elevation of mucosal tissue around the papilla makes snaring difficult. Therefore mucosal injection is not routinely recommended. Recently, “underwater ampullectomy” without submucosal lifting for lateral spreading tumor has been introduced, but its effectiveness is still under investigation .
33.4.2 Snaring and Transection
Electrosurgical snare resection is the most common technique. There is no specific type or size of snare for endoscopic ampullectomy. We usually use oval, braided polypectomy snares of 2–3 cm. However, in giant lesions (<3 cm), a large snare is difficult to handle under jejunoscopic view. The snare is placed with the tip on the oral side of the lesion. The snare is closed at the base, and the lesion is resected. In some reports, an incision is made with an electrosurgical needle knife circumferentially around the lesion to facilitate snare capture . When the lesion is held, the snare is stretched, and opposing tension has to be applied for snaring. Though there are no general recommendations regarding the optimal current and power output, there are reports of both pure cutting and blended cutting. Many endoscopists prefer the “blended” or “ERBE Endocut” mode, which aims to decrease bleeding by coagulation. We always use the “Endocut” mode with 100 W output power.
33.4.3 Retrieval of the Resected Tumor
Retrieval of the specimen is very important for the accurate evaluation and tumor staging. Immediately after the transection, the specimen is grasped by a snare and removed from the body in order to avoid intestinal migration. If the specimen is large, a basket catheter or a net forceps is useful. It is important that the tissue is not collected by aspiration through the endoscope, as this will cause the specimen to fragment, making it impossible to evaluate the cut end histopathologically.
Sphincterotomy is often performed after EP for facilitating pancreatic and biliary drainage. However, there is no consensus regarding sphincterotomy.
33.4.5 Treatment of Remnant Tissue
There is no standard technique for treating remnant tumor. Snares, biopsy forceps, and thermal ablation such as argon plasma coagulation are used for the treatment. Argon plasma coagulation is the most common and is useful for ablating remnant tumor as well as for hemostasis or prophylaxis of post-procedural bleeding. However it must be carefully applied to the tissue around the pancreatic and bile duct orifice because it may induce bile duct obstruction or pancreatitis by the thermal effect.
33.5 Complications of EP and Techniques for Their Prevention
Pancreatitis is the most common complication. Pancreatitis is considered to occur due to obstruction of pancreatic duct orifice and thermal damage to the pancreatic parenchyma. We sometimes encounter cases of pancreatitis in which bleeding induces obstruction of the pancreatic duct orifice or a pancreatic stent.
Pancreatic stent placement is recommended to prevent post-procedural pancreatitis. The aim of pancreatic stenting is to maintain the pancreatic duct orifice and to prevent pancreatic duct obstruction. One small randomized controlled trial concluded that pancreatic stenting prevented post-procedural pancreatitis . However, pancreatic damage by thermal ablation cannot be prevented. A recent retrospective study suggested that routine pancreatic stent placement may not be necessary. The diameters and lengths of pancreatic stents are not standardized. Generally, a pancreatic stent of 5Fr or 3Fr diameter is used after the resection of the tumor. It is sometimes difficult to find the pancreatic duct orifice. Some endoscopists reported that stent placement before the resection may reduce post-procedural pancreatitis . However once in situ the stent makes resection difficult and may interfere with en bloc resection. The technique of pancreatic duct wire-guided EP or retrieval of intraductally migrated pancreatic stents after EP has been introduced. A randomized controlled trial comparing wire-guided papillectomy and conventional papillectomy reported that there was no significant difference in the post-procedural pancreatitis or complete resection rates between the two methods [38, 39, 40, 41]. The removal time of stents is also not standardized. In our experience, we place a 5Fr pancreatic stent with a flap and remove it 7 days after the resection. In cases with pancreas divisum or dominant Santorini’s duct, discovered on evaluation by MRCP or EUS, a pancreatic stent is not placed.
Apart from the risk of post-procedural pancreatitis, obstructive cholangitis does not frequently occur except when caused by obstruction by a clot due to major bleeding. Therefore prophylactic biliary stent placement is generally unnecessary. Sphincterotomy is often performed, but this is not standardized. We do not add sphincterotomy, but a prophylactic biliary stent is placed to minimize the risk of cholangitis for cases in whom hemostasis was performed near bile duct orifice.
Post-procedural bleeding is one of the serious complications. The duodenal papilla is a hypervascular area. Antiplatelet agents and anticoagulated agents should be temporarily stopped as allowed by the condition of concomitant cardiovascular disease. Bleeding is often observed on the anal side of the resected margin . Some endoscopists perform argon plasma coagulation or semi-closure of scar for prevention of bleeding. However there is no data regarding additional procedures for the prevention of bleeding. Post-procedural bleeding can be treated by adrenaline injection, argon plasma coagulation, and clipping. Clipping is sometimes difficult with a duodenoscope because the existence of an elevator interferes with the opening of clips. When bleeding cannot be treated through hemostatic procedures, coiling by arteriography is helpful. Hemostatic procedures may induce perforation or pancreatitis. Therefore excessive hemostasis should be avoided.
Perforation usually occurs in the retroperitoneal area. The patients may not have peritoneal irritation signs. However, pancreatitis or bleeding is often observed concurrently with perforation. When perforation is suspected, evaluation by CT scan is informative. If perforation occurs, surgery may be considered, but selected patients can be treated conservatively with antibiotics and NPO [11, 13, 15].
33.5.5 Long-Term Results and Surveillance
The recurrence rate of ampullary adenoma after EP is reported to be 0–33% with a median follow-up period from 19 to 65 months [9, 10, 11, 12, 13, 16]. There is no consensus on the duration and modality of surveillance after EP. Recommendation of interval period varies; endoscopy with ERCP at 1, 3, 6, and 12 months after resection and then at yearly intervals for 5 years  and endoscopy at 3 or 12 months depending on the results of resection, 6, 12, 18, 24, and 36 months after resection , have all been suggested. In our institution, follow-up jejunoscopy and abdominal US is performed at 6-month intervals for 5 years for adenoma cases, and a yearly CT scan is added for patients with adenocarcinoma.
EP has been established as a first-line treatment for ampullary adenoma without ductal extension. There is still no consensus on preprocedural assessment, technique of EP, management of complications, or surveillance. Biopsy, EUS evaluation for large lesions, ERCP for further information, pancreatic stent placement for the prevention of pancreatitis, and endoscopic surveillance are recommended.
- 6.Suzuki K, Kantou U, Murakami Y. Two cases with ampullary cancer who underwent endoscopic excision. Prog Dig Endosc. 1983;23:236–9.Google Scholar