Endoscopic treatment of gastrointestinal perforations and leaks: Why, when, and how?

Abstract

Gastrointestinal leaks and perforations are severe conditions that require urgent management. While these conditions were traditionally treated surgically, a growing body of evidence suggests that endoscopic treatment may be more suitable in a significant proportion of cases. Furthermore, the evolution of therapeutic endoscopy has led to an increased number of such cases being diagnosed during the initial intervention, providing the opportunity for prompt endoscopic treatment. The decision between endoscopic and surgical therapy, as well as the specific type of endoscopic therapy to use, depends on the type and site of perforation, the timing, and the individual characteristics of each patient. It is essential that clinicians can accurately identify patients who are suitable for endoscopic treatment and those who require surgical intervention. Furthermore, there are many areas where high-quality data on the comparison of different modalities are lacking. This manuscript aims to illustrate the patients for whom endoscopic treatment may be indicated and the modalities that can be used for defect closure and prevention of complications.

Key Words: Perforations; Leaks; Endoscopy; Management; Clips; Modalities

Core Tip: Endoscopic management of perforations and leaks is becoming more and more prevalent. Additionally, new endoscopic modalities are constantly added to the gastroenterologist’s therapeutic arsenal, further reinforcing the concept of endoscopic closure as the initial treatment of choice in many cases. Gastroenterologists should be aware of current closure techniques and be able to determine which patients should be treated endoscopically and by which modality.

INTRODUCTION

Gastrointestinal (GI) perforations and postoperative leaks have long been one of the most common surgical emergencies, with their management carrying significant morbidity and mortality. However, with the emergence of interventional endoscopy, the management of these conditions has shifted toward the care of gastroenterologists, with the number of cases treated endoscopically increasing steadily[1,2]. Apart from the significant advances in endoscopic techniques and equipment, the comorbidities of many patients that present such findings as well as morbidity associated with surgical interventions, make endoscopic management the first choice for many cases. Furthermore, the significant rise in the utilization of interventional endoscopy has been accompanied by an increase in the incidence of perforations following such procedures, which necessitates that the gastroenterologist performing these procedures be capable of managing such complications[3,4].

The fundamental principles in managing GI wall defects involve bowel rest, intravenous fluid resuscitation, targeted antibiotic therapy, preservation of nutritional status, and continuous clinical monitoring. In addition, accurate localization of the defect through imaging, effective closure or diversion of the defect, and drainage of any associated fluid collections are essential components of care. Traditionally, the treatment of GI leaks and perforations relied heavily on conservative measures or surgical revision. Nonetheless, surgical procedures can be technically demanding and are frequently associated with a high risk of morbidity and mortality[5]. Over the past decades, there has been a growing shift toward endoscopic approaches as viable alternatives. Advances in therapeutic endoscopy have facilitated a significant transformation in the management of GI wall defects, shifting the focus from conventional surgery to minimally invasive endoscopic techniques[6]. A range of endoscopic modalities is now available to restore GI integrity, prevent further leakage and sepsis, facilitate drainage of collections, and ensure enteral nutrition. The choice of endoscopic modality depends on the clinical presentation, defect features (type of perforation, size, timing, and degree of contamination), as well as the availability of expertise[7].

The aim of this article is to highlight the subset of patients who can be treated endoscopically and to outline the modalities that can be used, depending on the characteristics of each case, thereby providing a practical guide for gastroenterologists and surgeons who encounter such cases.

METHODOLOGY/ SEARCH STRATEGY

A thorough search for relevant studies was conducted on PubMed, Google Scholar, and Cochrane Library through July 2025. Search terms included “gastrointestinal perforation”, “endoscopic closure”, “gastrointestinal leak”, “endoscopic vacuum therapy”, “over-the-scope clips”, “through-the-scope clips”, “endoscopic stents”, “tissue adhesives”, and “endoscopic suturing”. We also searched the references of selected studies and relevant reviews for unidentified studies.

TREATMENT SELECTION: TIMING AND DETERMINANT FACTORS

Acute perforation is defined as the detection of gas or luminal fluids outside the GI tract, or any endoscopically identified evidence of perforation observed during the procedure or in close temporal association with it[8]. Prompt diagnosis is crucial, as it facilitates endoscopic or surgical management in an uncontaminated field, thereby improving patient outcomes[9]. The management strategy for perforation is determined by several factors, including the timing of diagnosis, the presence and nature of luminal contents (clean vs contaminated), the perforation’s size and anatomical location, the patient’s overall clinical condition, the morbidity attached to the recommended method of surgical repair, the endoscopist’s level of expertise, and the availability of appropriate closure devices. Therapeutic approaches include endoscopic repair, conservative management, or surgical intervention. All aspects that should be taken into account when encountering acute perforation or leak are summarized in Figure 1.

Figure 1 Things to consider before choosing how to manage an acute perforation. GI: Gastrointestinal; OTSC: Over-the-scope clip; EVT: Endoscopic vacuum therapy.

Iatrogenic perforation during GI endoscopy should be promptly identified and followed by immediate attempts at closure, thereby minimizing its clinical implications. However, not all iatrogenic perforations are recognized at the time of endoscopy; therefore, clinicians should maintain a high index of suspicion when assessing post-endoscopy symptoms, particularly after high-risk interventions[10]. Early evaluation is warranted in the presence of atypical abdominal pain with distension, chest discomfort, subcutaneous emphysema, or dyspnea. Delayed presentation typically manifests with more severe clinical features, including systemic inflammatory response, hypotension, and altered mental status. European Society of Gastrointestinal Endoscopy recommends that, when an iatrogenic perforation is identified endoscopically, the endoscopist should document its size and location, provide photographic evidence, and record the endoscopic management undertaken. When the perforation is identified during the endoscopic procedure, completion of the planned intervention is recommended whenever feasible and clinically justified. Furthermore, in cases where clinical signs or symptoms raise suspicion of iatrogenic perforation following an endoscopic procedure, prompt and thorough evaluation with computed tomography is recommended[7]. Regarding GI leaks, both acute and chronic can result from inflammatory or malignant conditions; however, one of the leading causes is an anastomotic leak following GI surgery. Postoperative leaks remain a significant clinical challenge, contributing to considerable morbidity and potentially very high mortality rates - exceeding 60% when management is delayed[11]. Prompt recognition and timely intervention are crucial to improving survival rates. Clinical features suggestive of leakage of GI contents into the mediastinum, pleural cavity, or peritoneal space include fever, signs of systemic inflammatory response syndrome, septic shock, elevated C-reactive protein, and leukocytosis. The decision on how to treat GI leaks is based on the same factors as perforations, with the added significance of whether adequate percutaneous drainage and diversion of the luminal contents via a stoma are in place.

TREATMENT OPTIONS

A summary of all treatment modalities, along with their advantages and limitations, is presented in Table 1.

Table 1 Endoscopic modalities for gastrointestinal defect closure.

Endoscopic modality	Benefits	Limitations	Ideal use case
TTSC	Widely available and low-cost	Limited closure strength	Small defects
	Easy to deploy through standard endoscopes	Not suitable for large defects
	Short procedure time	Multiple clips may be required	Healthy tissue
	Multiple sizes	Spontaneous dislodgement
OTSC	Full-thickness closure and strong compression	Requires withdrawal and reinsertion	Large defects
	Durable closure with a single device	Difficult to remove	Fibrotic tissue
	Useful for larger defects		Area with easy access
Endoscopic stents	Early oral intake	Migration risk, especially SEPS	Esophagus
	Diversion of luminal contents, promoting healing	May cause pressure necrosis or ulceration
	Useful in leaks not amenable to direct closure	Need for repeat endoscopy for stent retrieval/replacement	Large defects
	Can cover large defects	Difficult use in lower GI
EVT	Highly effective for chronic leaks and large cavities	Multiple endoscopic sessions for sponge changes	Large cavities with abscess formation
	Promotes granulation tissue and healing	Patient discomfort	Patients with significant surgical risk
	Can be used when other modalities fail	Less widely available and technically demanding
Tissue adhesives	Minimally invasive and easy to apply	Limited efficacy as a stand-alone therapy	Small defects
	Useful as an adjunct to clips or stents	Variable durability	Combination with other modalities
		Limited data
Endoscopic suturing	Full-thickness, flexible, and customizable closure	Specialized equipment and advanced expertise	Large defects
			Fibrotic tissue
	Effective for larger defects	Longer procedure time	Area with easy access
	Can be combined with stents or other therapies	Limited availability and high cost	Post-ESD defects

TTSC: Through-the-scope clips; OTSC: Over-the-scope clips; GI: Gastrointestinal; EVT: Endoscopic vacuum therapy; SEPS: Self-expanding plastic stents; ESD: Endoscopic submuscoasl dissection.

Endoclips

Endoscopic clips represent the most commonly utilized and historically established modality for the closure of GI perforations and leaks[6]. Their initial application in clinical practice was reported in 1995 for the management of a postoperative anastomotic leak. Nowadays, the clinical indications for endoscopic clipping have significantly expanded to include not only GI perforations but also closure following natural orifice transluminal endoscopic surgery[12]. Additionally, endoclips are frequently employed as adjuncts to stent placement to minimize the risk of stent migration or dislodgement[13]. There are two main types of endoscopic clips available: Through-the-scope clips (TTSCs) and over-the-scope clips (OTSCs).

TTSCs: TTSC is introduced via the endoscope’s working channel and positioned to capture both margins of the defect effectively. Suction may be applied to approximate a greater amount of tissue, after which the assistant secures the clip by closing its jaws tightly. Once optimal alignment is achieved, the clip is deployed. TTSC systems are easy to use, and newer-generation clips offer improved maneuverability, including rotational capability and the option for reopening prior to release, allowing for enhanced precision. The clips can be deployed directly without the need to withdraw and reinsert the endoscope. TTSCs are particularly suitable for closing minor defects - typically under 1-2 cm in diameter - especially when the tissue margins are viable, non-everted, and well-approximated. However, their utility in managing larger or chronic defects is limited due to the relatively narrow clip arm width and reduced compressive force. This makes closure challenging when the tissue is inflamed, necrotic, or fibrotic. TTSCs are designed to detach spontaneously within 2-4 weeks[14], although premature dislodgement can lead to failure of defect closure and recurrence of GI wall disruption.

OTSCs: In cases of larger GI wall defects, the OTSC system is often the preferred therapeutic option. OTSC is a biocompatible nitinol device characterized by a bear-trap shape design, engineered to enable full-thickness closure of GI wall defects up to 2 cm in diameter. It is mounted on a transparent cap affixed to the distal end of the endoscope[15]. Among the most widely used OTSCs are the OTSC system (Ovesco Endoscopy AG, Tübingen, Germany) and the Padlock Clip (Aponos Medical Corp, Kingstone, New Hampshire). These clips are available in various sizes (mini, 11 mm, 12 mm, and 14 mm) and depths (3 mm and 6 mm), as well as tooth configurations tailored to specific clinical indications. There are three different tooth geometries for various clinical scenarios: Blunt teeth (type-a) for hemorrhagic lesions, teeth with small spikes (type-t) for thin-walled GI lumen, and elongated teeth with spikes (type-gc) for the thick stomach wall. The OTSC system is assembled by inserting a hand wheel into the rubber valve of the working channel and securing it to the endoscope with a Velcro strap. The applicator cap containing the clip is positioned at the distal end of the endoscope, with the release thread passed through the working channel - a mechanism similar to that used in variceal band ligation devices[15]. Tissue approximation is achieved either by circumferential suction or by employing a twin grasper to draw the defect edges together. Clip deployment is then accomplished by rotating the hand wheel. Due to its wider jaw span and enhanced compression force, the OTSC has been shown to effectively close gastric defects up to 20 mm and colonic defects up to 30 mm in diameter[16]. However, if closure fails or removal becomes necessary, extracting the OTSC can be technically demanding. Techniques such as argon plasma coagulation, bipolar cutting devices to weaken the clip structure, or cold saline infusion to increase metal pliability may be required for safe removal[17]. The Padlock Clip is a more recently developed OTSC system that differs in both design and deployment mechanism compared to the Ovesco clip. It features six inward-facing prongs that facilitate circumferential tissue approximation via 360° radial compression, enhancing full-thickness closure. The device comes preloaded in the open configuration and is mounted at the distal tip of the endoscope. Deployment is achieved through the Lock-It delivery system, which operates via a trigger wire aligned parallel to the scope and connected to an external handle. Preliminary studies have demonstrated the Padlock clip to be both safe and effective for managing GI wall defects[18]. Despite promising outcomes, clinical experience remains relatively limited, although a recent report has found it to have comparable efficacy to the OVESCO clip[19]. Therefore, further large-scale studies are warranted to validate its efficacy and define its optimal indications.

Endoscopic stent

Endoscopic stenting has become a well-established therapeutic option for GI defects, demonstrating high rates of both technical feasibility and clinical efficacy. It was initially utilized as a palliative measure for malignant conditions for nonsurgical candidates, particularly for obstructive esophageal neoplasms or tracheoesophageal fistulas. Over time, its indications were extended to include various benign GI conditions, such as anastomotic leaks and benign strictures. The primary objective of stent placement is to occlude the defect and redirect luminal contents, thereby facilitating mucosal healing. Additional benefits include the potential for earlier resumption of oral intake and a decreased likelihood of stricture development. A thorough pre-procedure endoscopic evaluation is essential to determine the defect’s extent, identify anatomical landmarks, choose the appropriate diameter and length of stent, and plan precise proximal and distal landing zones. In certain cases, additional drainage - either percutaneous or surgical - may be necessary[20]. Stent placement is typically performed under general anesthesia or conscious sedation, with fluoroscopic assistance. Following diagnostic endoscopy, a guidewire is advanced under direct visualization, and the stent is then deployed under fluoroscopic guidance. Complete expansion of the stent may take up to 24 hours to occur. In some instances, balloon dilatation is employed endoscopically to ensure complete radial expansion. To reduce the risk of stent migration, adjunctive fixation techniques such as endoscopic suturing (ES) or clipping have been utilized. These methods are most effective when applied at two or more opposing sites along the stent. Additionally, some endoscopists prefer to use slightly oversized stents to enhance stability[21]. Following deployment, two-plane fluoroscopic imaging is recommended to confirm adequate positioning and to serve as a baseline reference. While esophageal and duodenal defects are generally amenable to endoscopic stent placement, colonic defects are less frequently treated using this approach. Although current evidence on stent application in colonic leaks is limited, preliminary outcomes are promising, with a substantial proportion of patients demonstrating successful resolution of the defect. Nonetheless, stent migration and patient discomfort remain notable limitations[22].

Currently available stent options for GI defect management include self-expanding plastic stents (SEPS), self-expanding metal stents (SEMS) [either fully covered (FCSEMS) or partially covered (PCSEMS)], and, in some cases, vacuum-assisted closure (VAC) stents, although their clinical efficacy remains to be evaluated. Among the existing therapeutic tools, covered metallic stents - particularly FCSEMS - are most commonly utilized for the endoscopic treatment of upper GI wall defects[23]. FCSEMS are designed with flared ends to minimize the risk of migration, while radiopaque markers assist in fluoroscopic guidance during placement. Defect dimensions are first assessed endoscopically, followed by the insertion of a guidewire into the GI lumen to aid in accurate stent deployment.

SEPS: SEPSs have demonstrated safety, efficacy, and minimal invasiveness in the management of leaks and perforations, particularly in the esophagus[24]. The Polyflex stent (Boston Scientific, United States) is the most commonly used SEPS for this indication. Constructed from polyester and fully encased in silicone, it features a flared proximal end designed to minimize migration. Compared to SEMSs, SEPSs provide adequate sealing force due to the softer material, while the silicone covering prevents tissue ingrowth, thereby facilitating easier repositioning and removal. Nonetheless, SEPS placement presents certain limitations, such as a bulky delivery system and a relatively high migration rate[24]. Compared with conservative treatment, SEPS use is associated with earlier oral intake, shorter hospitalization, and lower in-hospital mortality[25]. Although the optimal timing for stent placement remains undefined, most studies recommend early insertion following diagnosis of a GI defect. Stent diameter is typically chosen based on the location of the defect: Cervical leaks generally require smaller calibers (18-23 mm) to reduce the risk of tracheal compression and foreign body sensation, while post-gastrectomy leaks are treated with slightly larger diameters (21-25 mm)[26]. After confirmation of healing, based on imaging with water-soluble contrast, endoscopic inspection, and clinical improvement, typically within 28 days, stents are generally removed. Stent migration is the most commonly reported complication, occurring in 8%-23% of cases in the short term and up to 40% after longer follow-up[24,26]. Strategies to reduce migration include selecting stents with larger diameters or anchoring the prosthesis to the GI wall using endoscopic clips.

SEMS: SEMS represent the most frequently employed stent type for managing upper GI perforations and leaks, according to an international expert survey on endoscopic management of upper GI leaks[27]. They are typically composed of either Elgiloy, which offers greater radial force and corrosion resistance, or Nitinol, which provides increased flexibility for angulated anatomy but lower radial force[28]. Only FCSEMS and PCSEMS are utilized for defect sealing and diversion, as uncovered stents cannot prevent leakage or redirect luminal contents. Clinical success rates for SEMS in the upper GI tract range from 48% to 100%[29], with optimal outcomes associated with early intervention - specifically, a short interval between surgery, diagnosis of the leak, and stent placement[30]. Factors linked to treatment failure include proximal cervical esophageal leaks, stent extension across the gastroesophageal junction, esophageal ruptures exceeding 6 cm, and anastomotic leaks with concomitant distal conduit defects[31]. The ideal stent indwelling time remains undefined; however, many studies suggest that approximately 30 days is sufficient for defect healing[31]. Reported adverse event (AE) rates for SEMS range from 20% to 72%, with stent-related mortality between 0% and 28% - lower than that observed after surgical repair (12%-50%)[32]. FCSEMS are associated with migration rates of up to 30%[33], although this can be reduced to approximately 15% with the use of ES, as demonstrated in a 2018 meta-analysis[21]. PCSEMS feature a membrane covering the stent body with uncovered flared ends that permit tissue ingrowth, thereby reducing the risk of migration. A pooled analysis of benign esophageal defect stenting reported an overall migration rate of 16.5%, with lower rates for PCSEMS (10.6%) compared to FCSEMS (21.8%)[32]. However, PCSEMS removal may be challenging due to tissue ingrowth. One proposed technique involves the placement of a SEPS within the PCSEMS for 6-10 weeks to induce pressure necrosis of the ingrown tissue, facilitating subsequent removal[34]. Swinnen et al[23] reported a 97.8% technical success rate for this SEPS-assisted retrieval method in 88 patients, with leak and perforation resolution in 77.6%[23], findings consistent with those of the Alazmi et al’s study[35]. Alternatively, argon plasma coagulation has been employed to ablate ingrowing tissue before extraction[23]. Stent migration can lead to additional AEs, including perforation or obstruction[36], and is often associated with altered anatomical configurations, the absence of luminal stenosis, and the inherently large diameter of the GI tract. Apart from migration, other stent-related AEs - including tissue overgrowth, bleeding, erosion, ulceration, perforation, reflux, and aspiration pneumonia - occurred at similar rates[37]. The application of SEMS has also been explored in the treatment of colorectal leaks and fistulas. However, the high motility and large luminal diameter of the colon contribute to stent migration rates approaching 40%[22]. The absence of commercially available covered enterocolonic SEMS further complicates management in these locations. Nevertheless, the off-label use of covered esophageal stents for enterocolonic fistulas has demonstrated encouraging outcomes[38]. In the lower GI tract, most studies report salvage success rates ranging from 50% to 100%, although patient-reported stent-related symptoms vary considerably. Migration remains the most frequent complication, occurring in 66.7% of cases, followed by anorectal pain in 58.3% and fecal incontinence in 25.0%. Despite these challenges, clinical success without the need for reoperation has been achieved in more than 83.3% of patients in the majority of published series, underscoring the potential role of SEMS in this setting[22,38]. When opting for SEMS therapy in the lower GI, several general principles should be observed: The exclusive use of fully covered SEMS is recommended, while stent placement should be avoided within 1 cm of the anal verge to minimize patient discomfort. Furthermore, any adjacent fluid collection should be drained beforehand, and the procedure should be avoided in the presence of active sepsis[39].

Vacuum-stent: The Vac-Stent^® (Microtech, Düsseldorf, Germany), a recently developed device, is a covered self-expanding stent with an internal lumen diameter of 12 mm, encased in a 5 cm polyurethane vacuum sponge layer connected to a suction catheter[40]. It has been introduced for the application of endoscopic vacuum therapy (EVT). The device is deployed transorally in a manner similar to conventional metallic stents, after which the suction catheter is redirected transnasally. Negative pressure is initially set at -125 mmHg and subsequently reduced to -75 mmHg. The negative pressure not only secures the stent in position and prevents dislocation, but more importantly, enables the combination of two therapeutic approaches: Negative pressure wound therapy and defect sealing via the stent[41]. The stent is typically removed after seven days, with repeat placement if leakage persists. Compared to traditional vacuum therapy, the VAC Stent eliminates many sources of patient discomfort and allows for oral food intake. Early prospective studies indicate high feasibility and efficacy in managing transmural defects of the upper GI tract. Additionally, initial case reports have documented its successful use in the colon, demonstrating promising outcomes[42]. However, further prospective studies are required to determine whether this device will assume a primary role in the management of larger leaks and complex anastomotic complications.

EVT

EVT represents a minimally invasive therapeutic modality for the management of anastomotic leaks. It is primarily employed to address the defect, utilizing a comprehensive endoscopic and radiological assessment that is essential for identifying and characterizing it, as well as for evaluating the size of any associated cavity. Larger defects, often accompanied by fluid collections, represent the most common indications for EVT. Although EVT applies to both acute and chronic GI defects, its predominant use is observed following rectal and esophageal surgery[43]. The Endo-SPONGE system consists of an open-pored polyurethane sponge connected to a suction catheter, which in turn is linked to a wound drainage system. The sponge may be tailored to the dimensions of the wound cavity to ensure optimal contact and efficacy. Following diagnostic endoscopy, an endoscope and overtube are introduced into the wound cavity to facilitate the precise placement of the sponge using a dedicated deployment device. Depending on the size and complexity of the defect, multiple sponges may be utilized sequentially during a single treatment session. The system delivers continuous, evenly distributed negative pressure across the tissue surfaces in contact with the sponge, thereby promoting adequate drainage and progressive reduction of the wound cavity. A significant limitation of this approach is the requirement for sponge exchanges every 48 hours to 72 hours until complete healing is confirmed by endoscopic and clinical assessment. The total duration of treatment may vary from a few days to several weeks, depending on the type of defect.

The primary aim of EVT in colorectal surgery is to facilitate early closure of diverting ileostomies and prevent the need for Hartmann’s procedure. EVT has demonstrated efficacy, safety, and good patient tolerability, especially when applied early in cases of distal anastomotic leaks in patients with a protective stoma and absence of systemic sepsis. Factors associated with EVT failure include delayed initiation of therapy, neoadjuvant treatment, lack of a protective stoma, age over 60 years, and male sex -many of which also constitute risk factors for anastomotic leakage post-surgery[44]. A retrospective analysis by Arezzo et al[45] evaluated the long-term efficacy of EVT in the treatment of colorectal anastomotic leaks in a cohort of 14 patients. The overall success rate was 79%, with a median treatment duration of 12.5 sessions (range: 4-40) and a median time to complete healing of 40.5 days[45]. EVT has also been extended as a viable therapeutic option for upper GI defects. Its application includes intra-cavitary placement for large defects or intraluminal placement for smaller lesions within the esophagus. A systematic review by Kuehn et al[46], encompassing over 200 patients with upper GI defects treated with EVT, reported a pooled success rate of 90% (range: 70%-100%), with low complication rates, including strictures (7.6%) and occasional bleeding after intra-cavitary sponge use[46]. Furthermore, a retrospective comparative study involving 71 patients assessed the outcomes of stent placement (SEMS or SEPS) vs EVT for nonsurgical closure of intrathoracic leaks. The EVT group demonstrated a significantly higher closure rate (84.4%) compared to the stent group (53.8%)[47]. Although randomized controlled trials are lacking, most retrospective studies have indicated EVT superiority over stent therapy in terms of success rates, mortality, and AEs[48,49]. The advantages of EVT compared to SEMS include continuous drainage of infected areas, the ability for repeated endoscopic evaluation of the defect, and applicability in all esophageal regions, including the cricopharyngeal area[50]. These findings suggest that EVT via Endo-SPONGE offers a highly effective and promising alternative approach to conventional stenting techniques for the treatment of anastomotic leaks.

Tissue adhesives and glue

Tissue sealants have been effectively employed in the management of anastomotic leaks and low-output fistulas, with reported success rates in the literature ranging from 55.7% to 96.8%[51]. The two most commonly used sealants are fibrin glue and cyanoacrylate. Fibrin glue is composed of two main components: Human fibrinogen reconstituted with aprotinin, and human thrombin reconstituted with calcium chloride. These components are delivered simultaneously via a double-lumen catheter, resulting in the formation of an absorbable, flexible fibrin matrix that mimics the initial stages of blood coagulation and wound healing. The adhesive properties of fibrin glue are optimized in relatively dry environments; thus, prior removal of purulent material and mucosal ablation around the defect are recommended before application[52]. In a retrospective series by Lippert et al[53] involving 52 patients treated with fibrin glue for GI leaks and fistulas, durable closure using fibrin glue alone was observed in 36.5% of patients, increasing to 55.7% when combined with additional endoscopic interventions such as cyanoacrylate, clips, or stents. The volume of fibrin glue administered ranged from 2 mL to 81 mL (median: 8.5 mL), delivered over 1 session to 40 sessions (median: 4 sessions). Nonetheless, surgical intervention was required in 23.1% of cases.

Cyanoacrylate (N-butyl-2-cyanoacrylate) is a synthetic adhesive that rapidly polymerizes upon contact with moisture, inducing local tissue necrosis and an inflammatory response that acts as a foreign body, thereby promoting tissue healing. Its strong adhesive properties remain effective even in the presence of gastric or pancreatic secretions. Additionally, cyanoacrylate exhibits antibacterial properties, making it suitable for use in infected areas[54]. Alpha-cyanoacrylate monomer has also been reported to achieve an 88% success rate in the closure of post-esophagectomy leaks[55]. However, the available evidence is heterogeneous, with variations in patient selection, indications, and outcome definitions, which makes interpretation challenging and introduces a risk of bias, impeding its incorporation as a strong first-line modality. A meta-analysis of 14 studies comprising 203 patients reported an overall closure success rate of 81% and a complication rate of 1%[56]. While these findings suggest that cyanoacrylate is a safe and effective option for leaks and fistulas, further prospective, controlled studies are needed to confirm its efficacy. Although tissue adhesives represent an increasingly important therapeutic tool, current evidence suggests their use is primarily limited to small, low-output defects. Furthermore, their most effective use may be as an adjunctive therapy in technically demanding closures, combined with other modalities. Well-designed prospective studies will be essential to determine their comparative efficacy and to evaluate potential benefits when combined with other modalities.

ES

Over the past two decades, multiple ES systems have been developed to achieve full-thickness closure of GI defects; however, most have demonstrated significant limitations, restricting their widespread clinical application. ES has become feasible with the development of the OverStitch device (Apollo Endosurgery, Austin, TX, United States), which enables single-operator surgical suturing through a flexible endoscope[57]. The original, single-use, disposable OverStitch device is mounted on a double-channel therapeutic endoscope (Olympus), facilitating full-thickness, interrupted or continuous suturing with both absorbable and non-absorbable material. Its main components include a needle driver handle, a cap-mounted suturing head, and an anchor exchange catheter, with adjunctive use of grasping forceps or a tissue helix device to enhance tissue apposition. A major advancement is the OverStitch SX, which can be mounted on single-channel endoscopes and is compatible with more than 20 endoscope models across four platforms. Despite these innovations, the system requires considerable expertise and formal training, limiting its use to tertiary care centers. Suturing is technically demanding in settings with limited luminal space or tangential defect orientation, and - as with surgical sutures - primary closure of large defects depends on robust, healthy tissue[58]. Nevertheless, there is limited evidence addressing its role in the primary closure of GI leaks and fistulas. In a multicenter retrospective study, Sharaiha et al[59] evaluated 122 patients undergoing ES, including 40 with fistulas (32.7%) and 15 with leaks (12.3%). Technical success rates were high, but sustained clinical success was achieved in 80% of fistulas and only 27% of leaks[59]. In a recent series of 20 postoperative leaks, Granata et al[60] adopted a multimodal, tissue-status-based strategy: (1) Primary ES in healthy tissue; (2) Combined suturing with FCSEMS placement and anchoring in compromised tissue; or (3) FCSEMS placement with anchoring when suturing was not feasible. Long-term clinical success was achieved in 80% of cases overall, with rates of 77%, 85%, and 75% for groups A, B, and C, respectively. AEs occurred in four patients, consisting of short distal esophageal strictures. Finally, a recent meta-analysis has evaluated three different modalities of ES in the setting of post-endoscopic submucosal dissection (ESD) defect closure, showing technical and clinical success rates of over 90% for all modalities[61].

In summary, ES offers a minimally invasive approach capable of proper full-thickness closure of GI defects, with encouraging technical success rates in the management of leaks and fistulas. Its use may be especially helpful in closing post-ESD defects, which can be exceptionally large. However, it remains a technically complex procedure requiring advanced operator skill, limiting its availability to specialized centers. While current evidence supports its feasibility, further prospective studies are warranted to clarify its long-term efficacy and safety.

TREATMENT IN SPECIFIC AREAS OF THE GI TRACT

Esophagus

Esophageal perforation is defined as a transmural disruption in the esophageal wall, permitting leakage of luminal contents into the mediastinum or peritoneal cavity and triggering local and/or systemic inflammatory responses. Although relatively rare, the esophagus is a common site of GI perforation, most frequently secondary to iatrogenic injury during endoscopic or laparoscopic procedures. Approximately half of perforations related to upper endoscopy occur in the esophagus. Diagnostic upper endoscopy carries a low risk of perforation (0.03%), with most events occurring in the esophagus[62]. In contrast, certain endoscopic interventions - such as dilation, mucosal or submucosal resection, septotomy for Zenker’s diverticulum, and foreign body extraction - are associated with a heightened risk of iatrogenic esophageal perforation[7]. Rare causes include diagnostic transesophageal echocardiography and esophageal biopsy[63]. Non-iatrogenic etiologies comprise spontaneous perforation in a diseased esophagus, Boerhaave syndrome, and injuries due to foreign bodies or external trauma. Reported 30-day mortality rates range from 12% to 30%, with early recognition and treatment within 24 hours representing the most critical prognostic determinant. The condition of the underlying esophagus is also an essential factor (presence of chronic inflammation, cancer, or acute disruption of the esophageal wall) with emergency surgical intervention associated with increased mortality[64].

Endoscopy alone has a limited role as a purely diagnostic procedure and should be performed in selected cases and undertaken with caution, as it carries the risk of enlarging the wall defect or disseminating contaminated fluids from the esophagus. In general, its use is justified when the clinical context suggests that therapeutic intervention can be performed during the same session or, following consultation with the surgical team, immediately before surgery. Anastomotic leak is a leading cause of postoperative morbidity and mortality after esophagectomy, contributing to prolonged hospitalization. Large retrospective series report leak rates of 8%-10%, with mortality ranging from 10% to 20%, influenced by surgical approach and patient comorbidities such as obesity, heart failure, vascular disease, diabetes, and renal insufficiency[65]. Patients requiring surgical management generally experience higher mortality compared to those managed non-surgically.

Conservative measures typically involve intravenous broad-spectrum antibiotics, nil by mouth, nasogastric decompression, pain control, gastric acid suppression, and hemodynamic support[66]. Percutaneous drainage is recommended when fluid collections are present, with samples sent for microbiological analysis. Cervical esophageal perforations are more amenable to non-surgical treatment due to reduced risk of mediastinal contamination. Successful management of esophageal iatrogenic perforations, whether conservative or endoscopic, depends on several key factors. Early recognition and intervention within 24 hours of perforation, a small defect size (< 1 cm for TTSCs, < 2 cm for OTSCs), a clean esophageal lumen with minimal or no contamination of the mediastinum, absence of significant comorbidities, and clinical stability post-perforation, all favor positive outcomes[67]. Surgical intervention is generally indicated in patients who are hemodynamically unstable, in cases of delayed diagnosis (more than 24 hours) with radiological evidence of free perforation or significant mediastinal/pleural fluid collections, in patients with comorbidities, or when local endoscopic expertise is unavailable or prior endoscopic treatment has failed[68]. Endoscopic options include defect closure with clips, stent placement to divert luminal contents, or EVT[69]. TTSCs have demonstrated success for closing esophageal perforations ranging from 3 mm to 25 mm (median 10 mm)[70], although their limited wingspan restricts their use in larger defects or in cases with fibrotic or inflamed margins due to delayed treatment or prior unsuccessful interventions. Larger perforations may be addressed with OTSCs, as demonstrated in a European multicenter cohort, where five esophageal perforations were successfully closed using OTSCs alone or in combination with TTSCs[71]. However, perforations of the esophagus present significant closure challenges due to the esophagus's tubular structure. Hagel et al[72] demonstrated that defects situated in the proximal or mid-esophagus, measuring over 20 mm in size, with ischemic or congested margins, or present for more than 72 hours, exhibit the lowest likelihood of successful closure with OTSCs[72]. Temporary stent placement, particularly SEMSs, is recommended for mid- and lower-esophageal perforations, with PCSEMSs preferred for defects spanning the gastroesophageal junction due to the higher migration risk of FCSEMSs[73]. Stents are particularly useful for large defects or perforations associated with malignancy, also providing symptomatic relief of dysphagia[74]. Assessment of complete healing following stent placement for iatrogenic esophageal perforation remains challenging, and the precise duration required for full tissue recovery is not yet established. Nonetheless, a stent dwell time of 4-6 weeks has been suggested to allow sufficient defect closure while minimizing stent-related complications[75]. For peroral endoscopic myotomy-related small mucosal flap perforations, TTSCs are usually sufficient and clinically inconsequential. Larger mucosal flap injuries with devitalized edges may be managed with TTSCs, ES, fibrin glue injection, and/or stenting[76]. Stenting is generally discouraged after peroral endoscopic myotomy due to the risks of erosion, migration, and inadequate sealing at the post-procedure esophagogastric junction; in severe cases, emergency esophagectomy has been reported[77]. EVT represents a novel approach for large or persistent esophageal perforations and leaks. The initial management of esophagogastric leaks is influenced by the presence of an extraluminal drain, which can mitigate mediastinal and/or peritoneal contamination. In the absence of a surgical drain, infection is nearly inevitable, highlighting the importance of drain placement prior to leak closure. The esophagogastric anastomosis following an Ivor-Lewis procedure represents one of the most frequent sites of postoperative leaks[65]. Early detection with minimal luminal content spillage allows for effective endoscopic isolation of the leak. For early postoperative esophagectomy leaks, all efforts should focus on primary closure using TTSCs, OTSCs, or ES, with additional placement of a fully covered stent. The SEMS should slightly exceed the native esophageal diameter to seal the defect without enlarging it, and stent migration can be prevented using endosuturing or an OTSC. Covered SEMS remain the most effective endoscopic intervention. OTSCs have been successfully employed in cases of small defects identified early in the postoperative period[73]. EVT is a valuable alternative for patients who do not respond adequately to SEMS. Recent prospective and multicenter studies have demonstrated promising outcomes for EVT in GI leaks, with reported closure rates ranging from 78% to 94%[78].

Stomach

Gastric perforations most frequently occur in the setting of peptic ulcer disease (PUD)[79]. Other non-iatrogenic causes consist of stomach cancer, foreign bodies, and caustic ingestion. Non-iatrogenic gastric perforations should be generally managed surgically due to the frequent presence of peritonitis and the difficulties in the endoscopic closure of unhealthy tissue margins in the stomach. Endoscopic management in non-iatrogenic perforations may be attempted in patients with significant comorbidities in whom surgery carries a very high risk of mortality. Even in this subset of patients, endoscopic management should be reserved for those with no signs of peritonitis and ideally within 24 hours of the perforation occurring.

Iatrogenic gastric perforation is a different story. The leading causes of perforation during endoscopy are modern polypectomy techniques, such as endoscopic mucosal resection (EMR) and ESD. The growing use of ESD for precancerous and cancerous lesions in the stomach has led to significant but gradually diminishing numbers of procedural perforations, as the procedure has become more and more common[80-82]. All perforations identified during endoscopic procedures should be initially managed endoscopically. Additionally, perforations identified within 24 hours of the initial endoscopy have been associated with good outcomes after endoscopic management[83]. While perforations identified more than 24 hours after the initial procedure have been traditionally treated surgically, studies have shown that selected patients can be treated endoscopically[84]. Nevertheless, the most significant factor to consider is the presence of signs of diffuse peritonitis or fluid collections on computed tomography, which would indicate the need for surgical treatment[7].

Regarding post-surgical gastric perforations, the presence of signs of diffuse peritonitis remains the primary factor in determining the course of treatment. The presence of postoperative drains may delay the emergence of peritonitis and enable endoscopic attempts at closure[85]. Furthermore, in the event that focal peritonitis with abscess formation is identified, endoscopic treatment of the perforation can be attempted in conjunction with percutaneous drainage of the abscess.

With regard to the modalities used to treat gastric perforations, the decision depends on the site and size of the perforation. In general, TTSC can be very effective in addressing defects ranging up to 1 cm, depending on the status of the stomach tissue (normal, fibrotic, inflamed, etc.)[6]. Usually, multiple TTSCs will be utilized, with closure commencing from the distal end of the defect and being aided by air insufflation to promote convergence of the defect borders[86]. For larger defects (1 cm to 3 cm), OTSCs are the preferred option, with rates of successful sustained endoscopic closure exceeding 85%[16,87]. Moreover, studies have shown favourable results in patients with acute perforation due to PUD, with similar mortality rates and shorter hospital stays compared to surgical management[88]. However, great care must be taken in the deployment of the OTSC, as a failed attempt may significantly impede the use of other modalities. The use of SEMS is restricted to perforations occurring at the site of strictures or leaks after bariatric surgery[89]. Finally, while ES presents significant promise as a potentially ideal method for gastric perforations and leaks, there is insufficient evidence (with few studies focusing mainly on the esophagus[60,90]) and limited availability and expertise to support its broad use as first-choice monotherapy for gastric perforations and leaks. Nevertheless, the combined use of clips and suturing may be justified in the majority of cases, as each gastric defect has unique characteristics, and endoscopists may need to employ creative schemes to achieve successful closure.

Duodenum (non-periampullary)

The vast majority of spontaneous perforations in the duodenum are due to PUD[91]. The anatomy of the duodenum, the underlying condition of the mucosa, and the timing of presentation (usually, endoscopic treatment is attempted within 12 hours of perforation) lead to a diminished role for endoscopic therapy in non-iatrogenic duodenal perforations. However, each case should be evaluated on its own merits, with patients having significant comorbidities and contained perforations without signs of diffuse peritonitis warranting an endoscopic attempt at closure, potentially combined with percutaneous drainage of fluid collections[92].

Iatrogenic perforations are usually related to advanced luminal or biliary endoscopic procedures. EMR and ESD in the duodenum confer a significantly higher risk of perforation compared to other parts of the GI tract[93]. Therefore, it is recommended that only endoscopists with significant experience, both in these procedures and in treating procedural perforations, perform them[7]. Furthermore, a major cause of duodenal perforations is Stapfer type I perforations, which are the second most common type of ERCP-related perforations[94]. As in all iatrogenic perforations, intraprocedural or early (< 12 hours) recognition of the perforation should prompt immediate endoscopic closure attempt. In this context, ESD- and EMR-related perforations may be easier to manage than ERCP-related, which can be challenging to locate and visualize both with a side-viewing and a forward-viewing scope.

When it comes to how to attempt endoscopic closure of duodenal perforation, TTSCs have been the traditional modality used for most cases of perforations < 1 cm, with the efficacy of closure reaching > 90%. OTSCs have been slowly gaining an advantage due to their ability to be used for larger defects and their significantly higher grasping power, with reports showing that patients treated with OTSCs exhibit high technical success and significantly shorter hospital stays compared to those undergoing surgery[95-97]. Their use is especially recommended for defects between 1 cm and 3 cm in diameter. The use of partially covered SEMS may be an intriguing option for Stapfer type 1 perforations, especially when combined with endoclips or sutures to stabilize their position[98]. However, potential biliary complications and low availability of such stents have hindered their emergence as a first-choice treatment.

Colon and rectum

Colorectal perforation typically occurs in the setting of colorectal neoplasms and with the use of modern polypectomy techniques. Nevertheless, there is also a small but not negligible risk of perforation in diagnostic colonoscopy (up to 0.16%[99]), with the sigmoid colon being the most prevalent site. The most significant factors influencing decision-making when treating a colorectal perforation are the site, timing, and bowel content. It is universally acknowledged that there is only a short window for endoscopic treatment, usually within a few hours of perforation. Significant luminal content and its free leakage in the peritoneal cavity essentially prohibit any endoscopic intervention. Furthermore, in cases where the perforation is acknowledged post-endoscopy, an easy-to-access site (e.g., rectum) favors endoscopic treatment and enables the use of more complex methods (OTSC, suturing, EVT in cases of leaks). Combining difficult access with an unclean colon greatly enhances the difficulty of closure, in which circumstances it is crucial to recognize the perforation during the procedure and attempt endoscopic closure with TTSCs.

Regarding the modalities that can be used, there are different options available depending on the presence of the factors as mentioned above. TTSCs remain the first choice for small perforations (< 1 cm), as well as for larger perforations in cases where it would take a significant amount of time to re-approach the site of perforation (e.g., cecum, right colon). OTSCs and suturing devices can be used in larger defects (greater than 1 cm) if the timing and bowel preparation permit it. EVT can also be used in substantial defects (> 3 cm), especially in patients unfit for surgery and in locations with easy access (mainly the rectum).

When it comes to leaks after colorectal surgery, the extent of luminal content spillage and the presence of an external drain are crucial for their endoscopic management. The presence of a drainage tube enables the use of clipping and suturing methods, with OTSCs being the most extensively studied. A recent pooled analysis of over 100 patients showed clinical success rates of greater than 75%[100]. In patients without adequate external drainage, EVT may be the most efficient treatment, with multiple studies demonstrating successful resolution of leaks in over 85% of cases[101-103]. Although numerous studies have demonstrated the efficacy of EVT in patients with concomitant external drainage, its prolonged treatment duration and requirement for multiple repeat endoscopies may ultimately render it less favorable than the OTSC in this subset of patients. It should be noted that most studies have analyzed EVT in the context of colorectal anastomosis, often with the concomitant presence of a stoma for fecal diversion[101]. The emergence of a dedicated colon Vac-Stent^® may be the solution enabling treatment of the leak without concomitant stoma creation, with a pilot study showing complete wound healing in 6 patients[104].

CONCLUSION

Recent advances in endoscopy have the potential to revolutionize the treatment of GI perforations and leaks. Gastroenterologists should be able to identify these conditions early and attempt endoscopic closure when it’s feasible. While most studies show excellent results with various modalities, this should not diminish the difficulty of managing GI defects, whether iatrogenic or not. Most published literature consists of case reports or small cohorts from centres of excellence, which may introduce significant publication bias. Therefore, gastroenterologists should endeavour to undergo substantial training in the use of modern techniques, such as OTSCs, suturing devices, and vacuum therapy, before using them in clinical practice. Finally, in whichever part of the GI lumen these defects are located, the most crucial aspect of management is recognizing which patients will benefit from endoscopic therapy and those for whom surgery is the only option.