Endoscopy 2022; 54(08): 745-746
DOI: 10.1055/a-1774-4831

Acute colonic diverticular haemorrhage: to band or clip?

Referring to Kobayashi K et al. p. 735–744
Department of Gastroenterology, Ulster Hospital, Belfast, BT16 1RH, United Kingdom
› Author Affiliations

There are several endoscopic options to stop acute hemorrhage from colonic diverticula, such as adrenaline injection, band ligation, endoloops, clips (both through the scope [TTS] and over the scope [OTS]), thermal therapy for example bipolar coagulation or argon plasma coagulation, and hemostatic powder. All have been reported to be effective in case reports and case series, although no head-to-head randomized controlled trials or systematic reviews of endoscopic therapies in this setting have been reported.

There have been recent guidelines from Europe and the USA that have evaluated the evidence and issued recommendations. The European Society of Gastrointestinal Endoscopy (ESGE) have suggested that either TTS/cap-mounted clip application or endoscopic band ligation (EBL) can be used as the preferred treatment [1]. A published review on treatment trends for colonic diverticular bleeding in Japan, which assessed five studies, concluded that EBL is superior to clipping in terms of the short- and long-term rebleeding rates, and fewer patients needing transcatheter arterial embolization or surgery [2]. There have however been reports suggesting that EBL has a risk of serious complications, such as delayed perforation, especially for right-sided lesions [3]. The British Society of Gastroenterology guidelines recommend the use of TTS clips as they can be applied without the need to remove the scope, are widely available and familiar, offer very high rates of immediate hemostasis, and can be used either alone or after adrenaline injection [4]. The American College of Gastroenterology guidelines also recommend TTS endoscopic clips as they may be safer than thermal therapy and applying them is easier than performing EBL, particularly for right-sided colon lesions [5]. In summary, if endoscopic therapy is required for acute colonic diverticular hemorrhage, two guidelines recommend clipping as being preferable to EBL, while the ESGE recommends that either option can be used.

“There may be technical reasons why band ligation may be preferable to clipping, for example when the precise bleeding point is unclear, as application of a clip needs to be precise for it to be most effective.”

The study by Kobayashi et al. published in this journal [6] is a multicenter cohort study from Japan that evaluates the effectiveness of endoscopic clipping versus EBL for colonic diverticular hemorrhage. Of 1679 patients with diverticular hemorrhage across 49 hospitals in Japan, 638 were treated with EBL and 1041 were treated with clips. In multivariate analysis, EBL was significantly associated with a reduced risk of early rebleeding (adjusted odds ratio [AOR] 0.46), late rebleeding (AOR 0.62), and need for interventional radiology (AOR 0.37), and with a shorter length of hospital stay (AOR 0.35). There were no differences in the rates of initial hemostasis, mortality, or need for surgery. Diverticulitis occurred in one patient (0.16 %) following EBL and two patients (0.19 %) following clipping. Perforation occurred in two patients (0.31 %) following EBL and none following clipping. These perforations occurred in the left-sided colon and both required surgery, but it is not stated whether the patients survived or not. Therefore, it was concluded that EBL appears to be more effective than clipping.

What are the strengths of this study? It is the first study comparing EBL with clipping that has a long follow-up, with a mean of 13.3 months. It has the largest number of patients with colonic diverticular hemorrhage treated with EBL and clipping reported to date in a single study. Previously such studies have been small, with a maximum of 790 cases. This is also a multicenter study, which suggests that the outcomes will be generalizable to multiple centers, not just a single center, where many such previously published studies have come from.

What are the limitations of this study? As this is not a randomized study, it is also not clear what the criteria for selection of patients for EBL versus clipping were. For example, was it up to the endoscopists preference or to certain anatomical characteristics? The characteristics of patients in the EBL group were significantly different to those of the clipping group in certain respects, raising concern that there may be a selection bias and that the study is not comparing like with like. Specifically, the EBL group had more alcohol drinkers, fewer hypotensive patients, and more with loss of consciousness, their hemoglobin levels were higher, fewer of them were on antiplatelets and steroids, more of them showed extravasation on contrast-enhanced computed tomography scanning, more received bowel preparation, and there was greater use of a distal attachment and water-jet scope. Such differences could potentially have altered the outcomes.

The authors do not provide information as to whether there were any differences in the duration of the procedure, as it is conceivable that the banding group would have undergone a longer procedure as the scope would have to be extubated to attach the banding device and then reintubated, and the duration might be even longer, especially if the diverticular bleed was in the right-sided colon. On this subject, there is no information on the site of the bleeding in the two groups and whether there were any differences in the proportion of right-sided colonic diverticula in each group, which is important as EBL may be associated with a higher perforation rate in the right-sided colon. It is not clear what the technical failure rates of these procedures were or if they were converted from one modality to the other.

We do not know what types of hospitals were involved (e. g. large teaching hospitals or community hospitals), as this may have an impact on how generalizable these results are to non-specialist endoscopists. It is not known if these results can be generalized to the Western population as this study, along with most other studies of EBL, was done in Japan.

There is also concern that the two perforations that occurred in the study were in the EBL group, previous groups having flagged the possibility of a higher risk of perforation with EBL. There were however no differences in the important end points of mortality and requirement for surgery.

In conclusion, does this study change our practice? There is insufficient evidence to-date to recommend EBL in preference to clipping for colonic diverticular hemorrhage. There may be technical reasons why EBL may be preferable to clipping, for example when the precise bleeding point is unclear, as application of a clip needs to be precise for it to be most effective. Some endoscopists may have a preference for EBL over clipping. EBL could be used as second-line therapy should clipping fail to achieve initial hemostasis. It would be helpful to see cohort studies, such as this study, performed in Western populations to see if the results can be replicated or, even better, randomized trials performed of EBL versus clipping.

Publication History

Article published online:
07 March 2022

© 2022. Thieme. All rights reserved.

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany