The NOTES experience highlighted the need for secure, full-thickness closure of the GI tract [4]. One response from industry was the development of large clips that are attached to the end of the endoscope (“over the scope”) and are designed for robust, full-thickness closures and apposition of even relatively diseased tissues. There are currently two of this type of clips in the market that function very similarly. The over-the-scope clip (OTSC) (Ovesco, Tübingen, Germany) is a nickel–titanium alloy (Nitinol) clip that is highly elastic and has shape memory [5] (Fig. 2). It is designed to cover lesions from 9–11 mm (depending on the size of the clip). The clips are designed with different forms (atraumatic, sharp, or blunt edges) in order to be used in a range of scenarios from iatrogenic perforations to chronic fistulas. The OTSC has an ability to grasp more tissue than standard endoclips creating full thickness closure of a gastrointestinal wall defect, even in the presence of inflammation or induration.

The OTSC clip is delivered with an applicator cap mounted on the tip of the scope. A thread is attached to the clip and it is released by tightening the thread with a hand wheel. The caps come in different sizes (11, 12, and 14) to fit different endoscopes and in two different depths (3 and 6 mm) for grasping more or less tissue. Tissue capture is achieved by both robust tissue retraction and by sucking tissue into the cap. Ovesco has also produced two types of tissue retraction devices as standard endoscopic graspers are seldom robust enough to adequately pull tissue like this into the cap. These include an “anchor catheter” and a novel twin grasper (Fig. 3). The “anchor” is a long flexible catheter with three retractable pins that pierce the tissue in order to retract it into the cap and assist in the placement of the clip. The twin grasper is similar to a traditional grasper, but has the ability to open the two jaws of the grasper independently. This feature allows the endoscopist to grasp one side of the defect, hold it without releasing, and then grasp the opposite edge and approximate the tissue before releasing the clip. The OTSC received ‘Conformite European’ (CE) certification in Europe in 2009 and 510(k) clearance by the FDA in 2010 [1].

The Padlock clip (Aponos Medical, Kingston, NH), is a similar design, but the firing mechanism is adjacent to the scope, and therefore, preserves use of the biopsy channel for instrumentation or suction/irrigation. It also comes in a variety of sizes and has a universal cap that fits on most types of endoscopes (Fig. 4). It too relies both on grasping the tissue and retracting and suction into the cap to achieve full-thickness closure. The Padlock received the FDA clearance in 2012, but is not currently cleared by a CE for sale in Europe.

There have been multiple case reports and several small series that have documented the ability of the OTSC to close acute perforations, leaks, and fistulas [2, 5–12]. Animal studies have shown the superiority of the OTSC versus regular endoscopic clips in closing iatrogenic perforations [13]. The success is higher in GI perforation than for fistulas [1], but the preliminary results are promising. In order to facilitate the endoscopic closure of chronic perforations and fistulas, cauterization, debridement, and curettage of the margins should be performed. These techniques might improve healing around the edges of the perforation. Of note, is that once these clips are deployed, their characteristics make it extremely difficult to remove them without damaging the tissue between their jaws. The endoscopist must be certain that the clip is being placed in the right position before deployment. It has been suggested that exposure to cold water might aid in the removal of a misplaced nitinol OTSC; however, in our experience, we have found this to be unreliable. The development of a retrieval mechanism or technique would be a desirable feature in these types of clips. Risks reported with the use of the OTSC are perforation, bleeding, early detachment, mucosal lacerations, and although it has not been reported, luminal obstruction is a hypothetical complication.

One of the strongest requests to industry from clinicians interested in NOTES was the ability to suture endoluminally. The result was a flurry of concepts designed that ranged from simple (T-tags [TAS, Ethicon, Blue Ash, OH]) (Figure 5) to more complex adaptations of laparoscopic suturing devices such as the Endostitch (Covidien, Norwalk, CN) Figure 6. There were even very complex flexible surgical platforms developed capable of bimanual suturing with standard surgical suture and techniques Fig. 7). Due to the costs of many of these technologies and the fact that they arrived during the onset of the global economic crises, only a few have succeeded in reaching the market.

A few reports exist to date showing the feasibility, safety, and durability of closures performed with the Overstitch. The first report in humans [20] presents a successful closure of a gastrocutaneous fistula after a percutaneous endoscopic gastrostomy (PEG) removal. The most common procedure the Overstitch is used for is bariatric revision surgery [21–26]. Thompson et al. reported on 25 patients who had weight regain after roux-en-Y gastric bypass due to pouch dilation or a patulous gastrojejunostomy [27]. At 1 year, patients lost an average of 11 kg with no adverse sequelae.

Covered self expanding stents represent an important tool for dealing with all forms of breaches in the GI tract. Stents available in the market are self-expandable metal and plastic stents. Metal stents are composed of various materials like stainless steel, nitinol (nickel-titanium) with or without a silicon membrane, and elgiloy (cobalt, nickel and chromium). Nitinol stents are very flexible and elgiloy stents are corrosion resistant and capable of generating high radial force [30]. These metallic stents may be partially or fully covered by a plastic membrane [30]. Depending on the brand, they may have a distal and/or proximal flare to prevent migration [30] (Fig. 12). Plastic stents are approved for benign or malignant esophageal strictures. They have a woven polyester skeleton that is fully covered with a silicone membrane [30]. Although metallic stents are designed for malignant obstruction, they are widely used “off label” in the treatment of GI tract perforations. Noncovered metallic stents are equivalent to covered metallic stents in terms of technical success, clinical success, long-term patency, and survival rates for the palliative treatment of malignant obstruction in the digestive tract [31]. While stent migration or slippage is uncommon when they are used for tumor palliation, it is a major problem when covered stents are used to treat perforations as they often have no narrowing to hold the stent in position. Stent migration is also more common in covered stents, but noncovered stents are more prone to tissue ingrowths and obviously would be unsuitable for perforation treatment [31]. Stainless steel stents might also migrate more often than nitinol ones [31]. A variety of techniques to prevent migration have been described, but none is totally perfect. On a systematic review on benign esophageal rupture or anastomotic leaks, van Boeckel et al. [32] reported no difference in clinical success between fully covered, partially covered metallic, and plastic stents. Stent migration (25 % of patients) was more common in fully covered stents (plastic or metallic) than in partially covered stents. Tissue ingrowth was not significantly different. They recommended leaving the stents in place for approximately 7 weeks to achieve healing.

A case report of a gastrointestinal-related fistula include a gastrocolonic fistula that recurred 4 months after placing an Amplatzer device which was finally closed with a Cardio-seal implant with no evidence of recurrence for 18 months [41]. Other clinical experiences include the use of these devices for closure of tracheoesophageal fistulas [43–46]. The experience has not been completely satisfactory. Coppola [43] reported the enlargement of a tracheoesophageal fistula and migration of the device into the broncial tree 2 months after placing it with subsequent occlusion success after using partially covered stents and Kadlec [44] reported a dislodgment of the disc 12 days after placement with enlargement of the fistula finally requiring operative repair. Lee obtained better results although the follow-up was short (1 month) and Repici [45] obtained the best results closing a gastro-tracheal fistula after a gastric pull-up for esophageal cancer. Several attempts were made to close the fistula, including surgery, and the final attempt with the septal occlusion device was successful showing no recurrence in 8 months of follow-up.

More chronic leaks can present with a wide spectrum of symptoms: once again, full-blown sepsis is a possibility, but also a more indolent presentation is possible. Sometimes, particularly with fistulous connections to other parts of the GI tract, the patient can be asymptomatic or have hard-to-diagnose conditions like low grade fevers, diarrhea, or weight regain after a bypass surgery. These patients require a high index of suspicion and a thorough workup. Contrast x-rays and/or computed tomography (CT) scans are usually the best mode of treatment though sometimes a diagnostic endoscopy is required for a definitive diagnosis. These patients are good candidates for a try at an endoscopic repair as surgery can be quite difficult and prone to complications.

Anastomotic Leak or Delayed Presentation Leak: Intermediate leaks, either ones that occur remote from surgery (from ischemic complications, ulcerations, etc.) or ones that were missed and presented later, require a more complex and multimodal approach for treatment. Primary closure is seldom possible or often requires adjunct treatments for success. A CT scan should usually be performed to document the extent of peritoneal, retroperitoneal, or thoracic contamination. Percutaneous drain placement should be established if indicated and possible. Endoscopy is the next step, with assessment of the breach. If the hole is found to have relatively fresh and pliable edges, primary closure can be considered. Standard clips are usually not feasible and either an OTSC or a suturing device is the better option.