Closing the lid on iatrogenic colonic perforations

Colorectal perforation is the most feared adverse event during colonoscopy. The perforation rate is generally <.1% after diagnostic colonoscopy but increases with therapeutic interventions, such as EMR (1.4%) and endoscopic submucosal dissection (ESD) (5.7%).^,  Prompt recognition is paramount to achieving a successful outcome and may allow for nonoperative management in select cases.

Several reports have described successful endoscopic closure of colonoscopy-related perforations using a variety of modalities, including through-the-scope (TTS) clips with or without an endoloop, over-the-scope (OTS) clips, band ligation, self-expandable metal stents, endoscopic suturing, or some combination thereof. However, evidence-based data demonstrating superiority of one technique over another are lacking.

In this issue of Gastrointestinal Endoscopy, Kantsevoy et al and Takamaru et al present retrospective outcomes data regarding the endoscopic management of iatrogenic colonic perforations with contrasting viewpoints on the optimal closure technique. In the small (n = 21), single-center study by Kantsevoy et al, iatrogenic colonic perforations were closed using either TTS clips (Resolution clip; Boston Scientific, Natick, Mass) or an endoscopic suturing device (OverStitch; Apollo Endosurgery, Austin, Tex) over a 6-year study period. Perforations ranged from 3 to 15 mm in size and occurred in various procedural settings, although the disproportionately smaller clip group (n = 5) encompassed only EMR-related perforations, whereas the suturing group (n = 16) contained mostly ESD-related perforations. After endoscopic closure, patients were referred for urgent surgical intervention if they demonstrated worsening abdominal pain or peritoneal signs. All patients in the clip group (mean perforation size, 3.4 mm; mean, 5.6 clips per patient; range, 2-9) developed increased postprocedural pain requiring urgent laparoscopic evaluation with subsequent colon resection (n = 4) or prompt repeat endoscopic evaluation with successful rescue endoscopic suturing of the defect (n = 1). Comparatively, only 2 of 16 patients (mean perforation size, 5.6 mm) in the endoscopic suturing group demonstrated notable postprocedural abdominal pain necessitating laparoscopic exploration, although neither patient required surgical resection. In contrast, the study by Takamaru et al presents supportive evidence for closure of ESD-related colonic perforations using TTS clips. This matched case-control study spanning 15 years included 24 patients with intraprocedural perforations and 240 control subjects among a large cohort of patients (n = 935) who underwent ESD for colorectal neoplasms. The mean perforation size was 5 mm, and defect closure required a median of 7 clips (range, 1-15). Technical and clinical success after clip closure of ESD perforations was achieved in 23 of 24 patients (96%), in stark contrast with the outcomes of the clip group in the study by Kantsevoy et al.

Most notably, the studies by Kantsevoy et al and Takamaru et al provide discrepant results regarding the utility of TTS clips for the management of iatrogenic colonic perforations (0% vs 96% clinical success rates, respectively) despite similarities in mean perforation size (3.4 mm vs 5 mm). Although no comparative closure technique is included in the report by Takamaru et al, Kantsevoy et al suggest that endoscopic suturing is superior to TTS clip placement for the treatment of full-thickness iatrogenic perforations. Although the study results are intriguing and hypothesis-generating, the 2 groups (clip vs suturing) in the report by Kantsevoy et al are not comparable in our opinion, and the study is grossly underpowered to determine a clinically significant difference.

Both studies raise additional issues that bear consideration. One of the important determinants in the selection of a particular technique for perforation closure is device availability. Unlike endoscopic clips, the OverStitch suturing system is a relative newcomer to the marketplace and has yet to achieve widespread adoption. Moreover, use of the suturing device involves a specific procedural cadence that is not immediately intuitive to novice users and is technically more complex than clip placement. Proficiency with the OverStitch device generally requires observation of procedures performed by experienced operators, training in experimental models, and, ultimately, proctor oversight during initial cases. The current iteration of the OverStitch suturing system can only be mounted on a specific double-channel upper endoscope (GIF-2T160 or GIF-2TH180; Olympus Corp., Tokyo, Japan), which may pose additional restrictions to its use, such as defect closure in the right side of the colon because of a lack of endoscope length and an inability to maneuver through constricted, fixed, or angulated segments of the colon (eg, diverticular-filled narrowed sigmoid colon). In this regard, the senior author and proceduralist in the Kantsevoy et al study, who is also a cofounder and shareholder of the OverStitch system, is to be congratulated for successfully suturing a sizable proportion of perforations on the right side of the colon (29% of cases) and in a shorter mean closure time compared with clip placement (13.4 vs 17.5 min, respectively). We surmise, however, that the expert-level suturing skills and excellent outcomes reported in the Kantsevoy et al study may not be generalizable to most practicing endoscopists. Thus, the authors’ conclusion that endoscopic suturing eliminates the need for rescue surgery should be tempered.

Newer generations of the OverStitch suturing system are currently under development, including mounting of the device on standard single-channel gastroscopes and colonoscopes, which may broaden its appeal and facilitate its use. Although we concur with Kantsevoy et al that endoscopic full-thickness suturing provides a more robust closure than TTS clip placement, the latter has been shown to be effective for small (subcentimeter) iatrogenic perforations that are recognized and closed immediately at the time of the procedure.^,  In both studies, the mean perforation size was approximately 5 mm, and the suboptimal outcomes in the clip group of the Kantsevoy et al study are the exception rather than the rule in comparison with published data. Ultimately, determination of the optimal technique for iatrogenic perforation closure in the colon would require prospective comparative assessment of TTS clip placement versus endoscopic suturing, although we realize that such a study may be difficult to perform given the relatively uncommon occurrence of this adverse event.

An endoscopic treatment alternative to TTS clip placement and suturing for perforation closure are the OTS clip devices, including the commercially available Padlock clip (Aponos Medical Corp, Kingston, NH) and the OTS clip (OTSC; Ovesco Endoscopy AG, Tubingen, Germany). Both OTS clip devices use cap-based technology, similar to a band ligator, to suction the defect and surrounding tissue before clip deployment, with or without the assistance of retraction devices, such as grasping forceps.^,  The OTSC device is more versatile than the Padlock clip, which is currently only compatible for use with a standard upper endoscope. The OTS clip devices can provide full-thickness closure with greater compression force and predictable durability than TTS clips but without the technical complexity associated with endoscopic suturing. Similar to endoscopic suturing, however, OTS clip procedures necessitate endoscope withdrawal for device loading, which may not be desirable in some circumstances and adds complexity to an already challenging situation. On the other hand, TTS clip placement is relatively straightforward and can be applied immediately after identification of an iatrogenic perforation, although closure of large (>1 cm) gaping perforations may not be as robust as that provided by endoscopic suturing or OTS clips.

Regardless of the modality used, a secure endoscopic closure is paramount in achieving an optimal patient outcome, minimizing the need for surgical intervention, and limiting additional medical costs. The cost associated with a successful endoscopic closure is quite favorable relative to the estimates of $14,000 to $17,000 for laparoscopic colon resection.^,  In the context of clinical practice in the United States, both endoscopic suturing and TTS clip placement (mean of 7 clips) would incur direct costs averaging $1000 for closure of iatrogenic perforations described in the 2 studies. Larger defects that require numerous clips will offset the cost savings of a TTS clip-based closure strategy when compared with alternative modalities. The OTS devices average $500 to $600 per clip and compare favorably with the aforementioned devices, with the ability to close larger defects (1.5-2 cm) with a single clip.

From a practical perspective, we believe TTS clip placement is a reasonable treatment option for closure of small (<1 cm) intraprocedurally recognized iatrogenic perforations with viable tissue margins but favor the use of a more robust closure method, such as endoscopic suturing or an OTS clip, for large gaping perforations that can be accessed by these devices. Suction of surrounding enteric contents, patient repositioning so the defect is in a nondependent position to minimize extraluminal egress of fluid, and use of CO₂ (instead of air) with minimum insufflation are important procedural steps to take once an iatrogenic colonic perforation is recognized. Rarely is the release of tension pneumoperitoneum via an angiocatheter or Veress needle required. It is also unlikely that most proceduralists will perform the air-leak bubble test, as described in the Kantsevoy et al study, which uses a decompression needle to assess the adequacy of perforation closure.

There appears to be a paradigm shift in the management of iatrogenic colonic perforations from surgical intervention to successful endoscopic repair in many cases. Although no ideal endoscopic closure technique currently exists that can be universally applied to all iatrogenic perforations, the proceduralist has access to several robust endoscopic closure devices that can provide an optimal outcome. Ultimately, the selection of a particular modality for perforation closure will be influenced by several factors, including size and location of the defect, device availability and limitations, operator skillset, and cost.