Luminal perforation after endoscopy is a dreaded complication that is associated with significant morbidity and mortality, longer and more costly hospitalization, and the specter of potential future litigation. The management of such perforations requires a multidisciplinary approach. Until recently, surgery was required. However, nowadays the endoscopist has a burgeoning armamentarium of devices and techniques that may obviate surgery. This article discusses the approach to endoscopic perforations in the esophagus and stomach.
Key points
- •
Endoscopic closure of perforations can be successfully achieved using a variety of devices.
- •
Even with endoscopic closure of perforations, the patient needs continued close monitoring with multidisciplinary input.
- •
Endoscopic closure of perforations effectively creates a leakproof seal, permits healing of perforation, prevents peritonitis, and limits peritoneal and mediastinal adhesions.
Introduction
Gastrointestinal luminal perforation may result from passage of the endoscope itself, diagnostic/therapeutic maneuvers (biopsy, dilation, ablation, polypectomy, endoscopic mucosal resection [EMR]/endoscopic submucosal dissection [ESD]), barotrauma (rare), or increasingly purposeful perforation, as in natural orifice transluminal endoscopic surgery (NOTES), including peroral endoscopic myotomy (POEM). Herein the authors emphasize endoscopic nonoperative management and their own and prior clinical experience in the literature, although animal studies are discussed regarding the rapidly evolving field of gastric perforation closure.
Introduction
Gastrointestinal luminal perforation may result from passage of the endoscope itself, diagnostic/therapeutic maneuvers (biopsy, dilation, ablation, polypectomy, endoscopic mucosal resection [EMR]/endoscopic submucosal dissection [ESD]), barotrauma (rare), or increasingly purposeful perforation, as in natural orifice transluminal endoscopic surgery (NOTES), including peroral endoscopic myotomy (POEM). Herein the authors emphasize endoscopic nonoperative management and their own and prior clinical experience in the literature, although animal studies are discussed regarding the rapidly evolving field of gastric perforation closure.
Epidemiology
A large retrospective analysis from the Mayo Clinic noted a perforation rate of 1 in 3000 upper endoscopy procedures, with a 38% higher rate with therapeutic intervention. Most of these perforations were esophageal (51%) and duodenal (32%), with only 3% gastric. The overall mortality was 17%, and 31% of deaths were related to esophageal perforations. Nearly half of these patients were treated nonoperatively; however, one-fifth of the patients failed nonoperative treatment within 3.7 days and required surgical intervention. Meta-analysis of 75 studies of esophageal perforation yielded a pooled mortality of 17% and a mean hospital stay of 33 days. The distal esophagus is the most common site in one review. Gastric ESD is used for submucosal lesion resection and noninvasive neoplasia; one group noted a 5% perforation for the latter indication.
General principles of management
There are several essential initial steps the endoscopist should consider after an actual or suspected upper gastrointestinal perforation. Most importantly, the endoscopist should remain calm because proper management may obviate surgical intervention. Prompt recognition of the perforation and patient positioning minimizes spillage outside the gastrointestinal tract and improves clinical outcomes. Insufflation should be switched to carbon dioxide if room air is used initially. Endoscopic assessment is paramount, with attention to size, edge characteristics, and bleeding. There are a variety of potential closure methods. Parenteral antibiotics and proton-pump inhibitors should be administered immediately and attention given to cardiopulmonary issues. Endotracheal intubation is advisable, although it should not delay closure attempt in an otherwise stable patient. Surgery (general/thoracic) should be notified and the patient triaged for intensive care unit monitoring.
Judicious use of radiologic examinations may be helpful after perforation, including after attempted closure. A lateral neck film may detect air after cervical esophageal perforation. Posterior and lateral chest and abdominal radiographs are ordered for more distal esophageal and gastric perforations. Contrast studies (eg, contrast esophagography to assess for esophageal perforation) using water-soluble agents (eg, gastrograffin) are valuable for diagnosing perforation, although occasionally perforations are missed because of technical considerations. Contrast may also be injected during endoscopy to assess perforation and degree of closure after intervention. Computed tomography (CT) is useful when traditional contrast studies are negative, and directs interventions such as drainage of collections. CT can demonstrate mediastinal and subcutaneous emphysema, pleural effusion, and pneumothorax. Patients with conservative treatment of perforations require vigilant clinical and radiographic follow-up.
The mere presence of extraluminal air does not mandate surgery, nor is the amount proportionate to perforation size. On the other hand, air can dissect through tissue planes, and a compartment syndrome caused by air pressure is an emergency. Therefore, appropriate needles such as Veress needles or small-caliber angiocaths should be available to decompress tension pneumothorax and pneumoperitoneum. Needle insertion into the abdomen must be away from areas of potential organ injury (eg, away from surgical scars), and time must be allowed for adequate decompression. Failure to close an esophageal perforation usually mandates surgery. The stomach, on the other hand, is more “forgiving,” and small gastric perforations may be amenable to conservative management. Gastric perforations usually benefit from nasogastric tube insertion for decompression and diversion of luminal contents. In some cases, nonoperative management may be considered in conjunction with interventional radiologic drainage. Patients treated conservatively should keep fasting (nothing by mouth) with intravenous fluids and analgesia. Parenteral nutrition is considered, but selected patients should preferably receive enteral tube feedings distal to the pylorus if technically possible.
Endoscopic closure of esophageal, gastric, and colonic perforations, the subject of this review, has been investigated extensively in animal models. A recent review of these data concluded that the initial pilot studies and subsequent randomized controlled studies comparing endoscopic closure with surgical closure have demonstrated that “endoscopic closure of esophageal, gastric, and colon perforations is technically feasible, effective in creating a leak-proof closure, permits healing of perforation, prevents peritonitis, and limits peritoneal and mediastinal adhesions.” This review focuses mainly on clinical data from human studies pertaining to esophageal and gastric perforation closure.
Esophageal perforations
General Considerations
Esophageal perforation is encountered more commonly with therapeutic interventions such as stricture dilation, dilation for achalasia or eosinophilic esophagitis, foreign body extraction, or after ESD, but may also occur with diagnostic procedures such as a challenging esophageal intubation. Symptoms relate to site and extent of perforation and spillage of intraluminal contents. Subcutaneous emphysema is common. Neck pain is seen with cervical perforations, whereas severe chest pain, often with emesis, is noted with distal esophageal perforations. Fever, chills, tachycardia, and hypotension occur with development of sepsis.
Kuppusamy and colleagues have demonstrated the evolution of esophageal perforation management over the past 20 years from surgery to nonoperative methods such as endoscopy and/or interventional radiology placed drains. Small esophageal perforations (<2 cm) can be closed with clips. Larger perforations usually require a covered self-expanding metal stent (SEMS). Perforations associated with distal obstruction arising from benign or malignant strictures or other conditions such as achalasia are also best served with a stent to alleviate high luminal pressures in the area of the perforation that may result in failure of endoscopic closure. Perforations with everted edges are especially challenging, and may be best treated with endoluminal suturing or large over-the-scope clips (OTSCs), although use of these devices can be challenging in the esophagus, especially the proximal esophagus, owing to space constraints. Cervical esophageal perforations can often be treated conservatively because of containment by the neck fascial structures. This circumstance is fortunate because endoscopic closure of any type is difficult in the narrow tapered cervical esophagus, and patients are less tolerant of proximally deployed stents near the upper esophageal sphincter.
Clips
Through-the-scope (TTS) endoscopic mucosal clips can be used to close small perforations provided that the tissue surrounding the edges is compliant and viable. TTS clips are made by several companies and differ in terms of their diameter, rotation capacity, and deployment capability after multiple openings ( Fig. 1 ). Table 1 compares the key features of the more commonly used TTS clips. Occasionally they can be used to close larger, nongaping perforations. The endoscopist and assistant should be well versed with their chosen device, and coordinate their actions. Raju recommends approaching a linear tear from above, with the first clip tenting up tissue near the top edge of the perforation, and approaching a circular perforation from below with intended transverse closure. In either instance, multiple clips are applied in a parallel fashion to achieve closure. Enhancing maneuvers include orienting the clip in the intended position and simultaneously applying suction to invert and approximate the perforation edges, and only deploying the clip when the edges are approximated. A double-channel endoscope can be also used, which allows for a long grasping forceps to be passed through one channel to grasp the edges and initiate the closure. Then a mucosal or TTS clip is used through the other channel parallel to the grasped tissue, alongside the grasper, with subsequent deployment of the clip.
| Characteristics | QuickClip2 QuickClip Pro (Olympus) | Resolution Clip (Boston Scientific) | Instinct Clip (Cook Medical) |
|---|---|---|---|
| Sheath diameter (F) | 7 7 | 7 | 7 |
| Sheath length (cm) | 235 230 | 230 | 230 |
| Jaw opening width (mm) | 9, 11 11 | 11 | 16 |
| Reopening and repositioning ability | No Yes | Yes | Yes |
| Rotating ability | Yes Yes | No | Yes |
| Clip material | Stainless steel Elgiloy | Stainless steel | Stainless steel and nitinol |
| Deployment | Two-step Three-step | Three step | Two step |
| MRI compatible | No Up to 3 T | Up to 3 T | Up to 3 T |
| Approximate cost ($, per clip) | 65 Not available | 170 | 150 |
There is limited literature supporting the use of TTS clips in the closure of esophageal perforations, and several series include noniatrogenic perforations in addition to leaks and fistulas. Table 2 summarizes published case reports and case series describing successful esophageal perforation closure with endoscopic mucosal clips. TTS closure has been shown to be successful in a variety of circumstances including EMR, Boerhaave syndrome, and balloon-, tube-, and endoscope-related perforation, without the need for prolonged hospitalization.
| TTS Clip Closure of Esophageal Perforations | Type of Study | Comments |
|---|---|---|
| Shimizu et al, 2004 | Case series | 3 patients after EMR: successfully closed |
| Qadeer et al, 2007 | Case report and pooled analysis | Pooled analysis: 11 articles with 17 patients: median healing time = 18 d Success diminished with chronicity |
| Gerke et al, 2007 | Case report | Proximal perforation caused by diagnostic EGD. Successful closure |
| Fischer et al, 2007 | Case series | 2 diagnostic, 1 dilation, 1 laser. All successfully closed |
| Mangiavillano et al, 2010 | Pooled analysis | 10 patients: all successfully closed |
| Rokszin et al, 2011 | Case report | Boerhaave syndrome: no surgery required |
| Jung et al, 2011 | Case report | Blakemore tube insertion complication: no surgery required |
| Coda et al, 2012 | Case report | After achalasia balloon dilation: no surgery required |
| Huang et al, 2014 | Case series | Esophageal perforation caused by duodenoscope: all successfully closed |
An OTSC (Ovesco, Tübingen, Germany) was developed for closure of mural defects up to 18 mm in size and bleeding ulcers ( Fig. 2 ). The device comes in various sizes, includes a large-caliber nitinol clip that fits over a cap at the tip of the endoscope, and is deployed in similar fashion to bands from a variceal ligator. The OTSC has been successfully deployed throughout the gastrointestinal tract. Table 3 summarizes published case reports and case series describing successful esophageal perforation closure with OTSCs. In a European multicenter prospective cohort study, all 5 patients with esophageal perforations had successful closure with the OTSC. Ironically, in this study 1 patient with a duodenal perforation sustained an esophageal perforation with insertion of the device. In other series, 3 patients with esophageal perforation had successful OTSC closure. There has also been mention of successful closure of esophageal perforation caused by nasogastric tube placement with OTSC. Hagel and colleagues reviewed the predictors of successful use of the OTSC, and found that perforations located in the proximal and mid esophagus, greater than 20 mm in size, with ischemic or inflamed margins or present for greater than 72 hours, were least amenable to OTSC closure.
| OTSC Closure of Esophageal Perforations | N | Technical/Clinical Success (%) | Other Treatments | Complications |
|---|---|---|---|---|
| Kirschniak et al, 2011 | 7 | 100/100 | — | — |
| Baron et al, 2012 | 1 | 100/100 | — | — |
| Voermans et al, 2012 | 5 | 100/100 | Adjunctive clips and second OTSC | — |
| Hagel et al, 2012 | 4 | 25 (1/4) | Surgical suture | Enlargement of perforation |
| Haito-Chavez et al, 2014 | 10 | 100/100 | — | — |
| Jacobsen et al, 2012 | 3 | 100/100 | — | — |
| Nishiyama et al, 2013 | 1 | 100/100 | ||
| Ramhamadany et al, 2013 | 1 | 100/100 | — | — |
| Ferreira et al, 2013 | 1 | 100/100 | — | — |
| Bona et al, 2014 | 1 | 100/100 | — | — |
Stents
As already mentioned, endoscopic esophageal stenting is an effective means of closing larger esophageal perforations because it minimizes unfavorable pressure gradients related to strictures and achalasia that may cause an endoscopic closure to fail. Care should be taken to center the stent after deployment on the perforation site. Stents are less optimal in the proximal esophagus (foreign body sensation) and with a previously normal gastroesophageal junction (gastroesophageal reflux disease and migration). There had been some initial work with a plastic expandable stent (Polyflex; Boston Scientific, Natick, MA, USA), which was found to have unacceptably high migration rates. Most recent published experience has been predominantly with covered SEMS. These stents are made by several different companies and come in a variety of dimensions, radial forces, and mechanisms of deployment and removal. There is also a biodegradable stent, although it is not currently available in the United States (Ella-CS, Hradec Králové, Czech Republic).
Table 4 summarizes published series describing successful esophageal perforation closure with stents. Some have advocated smaller-diameter stents for more proximal perforations. There are insufficient data to suggest superiority of any particular expandable metal stent. Optimally, they are inserted with both endoscopic and fluoroscopic guidance. Migration represents a significant problem and is one of the main causes of failure. Attempts to secure these stents against migration have included fixation of the stents with clips, endoscopic sutures, or even a bridle, the latter 2 methods probably representing more reliable fixation than can be achieved with TTS clips.
| Study | No. of Perforations | Stent | Migration (%) | Mortality (%) | Clinical Success (%) |
|---|---|---|---|---|---|
| Siersema et al, 2003 | 8 | Flamingo/Ultraflex | 9 | 0 | 93 |
| Gelbmann et al, 2004 | 4 | Polyflex | 30 | 33 | 66 |
| Johnsson et al, 2005 | 20 | Ultraflex | 14 | 23 | 77 |
| Fischer et al, 2006 | 15 | Ultraflex/Niti S | — | 7 | 93 |
| Ott et al, 2007 | 4 | Polyflex | 33 | 50 | 50 |
| Freeman et al, 2007 | 17 | Polyflex/Wallflex | 18 | 0 | 94 |
| Tuebergen et al, 2008 | 10 | Ultraflex | 6 | 15 | 81 |
| Kim et al, 2008 | 9 | Salivary bypass | 35 | 11 | 77 |
| Leers et al, 2008 | 9 | Ultraflex | 3 | 0 | 100 |
| Kiev et al, 2007 | 9 | Polyflex | 21 | 0 | 100 |
| Salminen et al, 2009 | 8 | Ultraflex | 10 | 30 | 70 |
| van Heel et al, 2010 | 31 | Ultraflex/Polyflex | 33 | 15 | 74 |
| Kuppusamy et al, 2011 | 52 | Polyflex/Wallflex/Niti S | 21 | 0 | 100 |
| Dai et al, 2011 | 6 | Polyflex | 35 | 17 | 83 |
Related posts:
Comprehensive Review of Adverse Events Associated with GI Endoscopy
Minimizing, Recognizing, and Managing Endoscopic Adverse Events
Complications of Enteroscopy: How to Avoid Them and Manage Them When They Arise
Endoscopic Retrograde Cholangiopancreatography–Related Adverse Events
Endoscopic Retrograde Cholangiopancreatography (ERCP)-Related Adverse Events
Closing Perforations and Postperforation Management in Endoscopy
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
