Endoscopy 2022; 54(08): 795-796
DOI: 10.1055/a-1762-5625

Preemptive clipping for post-ampullectomy bleeding: the jury is still out

Referring to Park SW et al. p. 787–794
Lars Aabakken
Department of Transplantation Medicine, Oslo University Hospital-Rikshospitalet, Oslo, Norway
› Author Affiliations

Endoscopic resection of nonmalignant ampullary tumors is among the significant utilities of endoscopic retrograde cholangiopancreatography (ERCP). Even in lesions that are borderline resectable based on computed tomography or endosonography, an endoscopic ampullary resection offers vital histology sampling to inform further treatment and follow-up.

The utility and technical details of endoscopic ampullectomy are commendably described in a recent European Society of Gastrointestinal Endoscopy guideline [1]. The guideline also describes the potential complications of the procedure, which, in addition to the recognized risk of procedure-related pancreatitis, also include cholangitis, perforation, and bleeding. Bleeding may occur perioperatively, and therefore mandates a period of observation after snaring, but this is mostly manageable endoscopically. However, the significant risk of post-procedural (delayed) bleeding is a bigger concern, partly because of patient delay and partly because the subsequent endoscopic therapy may be complicated by intraduodenal clots, as well as the deteriorating condition of the patient.

The risk of delayed bleeding after resection is well recognized in ampullary [2] and nonampullary adenomas in the duodenum [3]. However, the resection of ampullary lesions presents at least two additional issues compared with nonampullary adenomas (and indeed other polypectomy situations). First, the resection site is in direct connection with the pancreatic duct orifice, posing a significant risk of pancreatitis regardless of endoscopic maneuvers. Moreover, hemostatic clipping [4], which is otherwise a vital tool in endoscopic hemostasis, is significantly hampered by use of the duodenoscope owing to the mechanical effects of the elevator, the use of which is needed to appropriately target the bleeding point.

“The efficacy of clipping for bleeding prevention was not proven in this study, even though there was a tendency toward fewer delayed bleeds. However, the number of cases of pancreatitis was similarly different but in the opposite direction, raising the question of whether clipping simply replaces one complication with another.”

Thus, the study published in this issue of Endoscopy by Park et al. [5], represents a very logical assessment of preemptive hemostatic clipping of the ampullary resection site, using a potentially more effective hemostatic clip [6]. This clip purportedly has a more flexible mechanical construction, offering improved functionality even when used through a duodenoscope. The authors randomized 78 consecutive patients with an indication for endoscopic ampullectomy to either standard treatment (without any clipping) or standard plus preemptive clipping at the caudal end of the resection site, using this novel clip design. The primary outcome was the incidence of delayed bleeding > 4 hours post-procedure, while also looking at other pertinent parameters, particularly other complications such as pancreatitis.

Technical success in clip deployment was 100 %. Delayed bleeding was seen in 31.6 % and 15.0 % in the nonclipped and clipped groups, respectively, a numerical though nonstatistical difference. However, the incidence of post-ERCP pancreatitis was in favor of nonclipping (5.3 % vs. 17.5 %), but again without reaching statistical difference. No other differences were demonstrated between the groups.

So, do these results suggest routine use of caudal margin clipping of post-ampullectomy mucosal defects? Though technically surely feasible, there are several questions that need to be addressed before universally adopting this strategy, questions that have also been partially raised by the authors of the paper.

The efficacy of clipping was not proven in this study, even though there was a tendency toward fewer delayed bleeds. However, the number of cases of pancreatitis was similarly different but in the opposite direction, raising the question of whether clipping simply replaces one complication with another. Furthermore, all bleeds were mild or moderate in this study. This compares well to the general idea that late bleeds are rarely catastrophic, whereas pancreatitis may indeed be. Results from this randomized trial were not as convincing as those of a similar study from Japan [7], emphasizing the need for properly designed randomized trials such as this one by Park et al.

The understandable urge to avoid ductal injury with the clip seems to be the main argument for placing the clip at the caudal end of the cut surface. However, this may not be the location most likely to bleed, given the relative location of vasculature around the papilla [8]. This may reduce the overall efficacy of the method, but will be difficult to remedy. Other hemostatic modalities may conceivably be better options and should definitely be compared with clipping before recommending a change of practice. Variants of hemostatic powder are still being evaluated for their optimal role in preventing delayed bleeding after resections [9] [10], but conceivably, the ampullectomy site might be a good indication for preemptive spraying, similarly to spraying the defects created by endoscopic mucosal resection or endoscopic submucosal dissection, potentially avoiding the risks associated with periampullary clipping.

However, the utility of a clip that is better suited for use with a duodenoscope should not be underestimated. Not uncommonly, acute duodenal bleeds are best addressed by the duodenoscope, and better clipping options in this situation would be very welcome. Other indications for clipping with a duodenoscope in the duodenum are also easily conceivable, including resection defects, fistulas/perforations, and cannulation of selected intradiverticular papillae.

Thus, although the role for preemptive post-ampullectomy clipping remains to be defined, and compared with alternative modalities, the introduction of a better clip to our therapeutic armamentarium surely is a valuable improvement.

Publication History

Article published online:
18 February 2022

© 2022. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany