Management of Esophageal Perforation After Therapeutic Endoscopy

Iatrogenic injuries account for up to 60% of all cases of esophageal perforation.¹ The risk of perforation increases significantly from 0.6% for purely diagnostic endoscopy to 6% for operative procedures.²

There are three main methods for reducing the burden of esophageal perforation due to operative endoscopy: improving training, particularly for advanced therapeutic procedures; respecting the indications of each procedure for each patient; and demonstrating prompt recognition of complications and their early and appropriate management. The interesting case report by Petersen³ focused on the last method by demonstrating the successful management of a linear perforation that developed after pneumatic dilation of postradiation esophageal stricture, by using a self-expanding plastic stent.

Management of perforations traditionally involves surgery or an aggressive approach, particularly for patients with significant mediastinal contamination who are not suitable for conservative management. Despite meticulous surgical techniques, a reduction in operating time, adequate antibiotic therapies, and advances in anesthetic management and postoperative care, the incidence of postoperative complications for these patients remains high.

Because many endoscopic perforations are small, well-defined, and have limited contamination, both the repair of the perforation and the diversion of luminal contents can now be accomplished via endoluminal means, which allows for the avoidance of surgery and its morbidity. Unfortunately, clinical evidence regarding the endoluminal closure of perforations is limited to case reports and case series; there have been no randomized controlled clinical studies in this area.

Largely based upon our understanding of the principles of surgical management, there are 3 important goals when treating patients with esophageal perforation: preventing ongoing soilage; providing debridement of devitalized tissue; and performing wide drainage.⁴

Choosing a therapeutic option for an esophageal perforation also requires differentiating between acute and chronic cases of perforation, as they are distinct clinical entities. Acute perforations are potentially life-threatening emergencies in which prompt closure is required to eliminate contamination of visceral spaces. In contrast, chronic perforations are smoldering problems complicated by abscesses and fistulas. In addition, damaged tissues have different susceptibilities to endoscopic handling. A vital and elastic tissue, just like new lesions, can easily be clipped or sutured. On the other hand, in older lesions, where the tissue is friable, necrotic, or callous, it is often easier to promote closure via secondary intention.

Currently, endoscopic clips are the only devices available on the market for closure of perforations. Suturing and stapling devices are not yet available for clinical use. Endoclips may be adequate and may signif-cantly reduce the need for surgery, particularly for the closure of linear or regular esophageal perforations from several millimeters to 2 cm in size. Multiple clips and/or multiple separate endoscopic sessions are required to close larger perforations.⁵

For cases of irregular perforations or deep-penetrating lacerations of the esophageal wall, a powerful nitinol over-the-scope clipping system (OTSC system, Ovesco Endoscopy) was developed to ensure the full-thickness approximation of the edges. This device has also been used to successfully close gastrointestinal leaks and bleeding lesions refractory to standard treatment.⁶

Sewing devices are in the preclinical phase of evaluation. Experience with endoscopic suturing of esophageal perforations and fistulas is limited or nonexistent.

Esophageal stent insertion has been shown to be successful in the closure of acute esophageal perforation immediately after its detection and in the closure of longstanding perforations in patients who are not candidates for surgery.⁷

Stents may be a better option in perforations or fistulas larger than 2 cm, in defects with everted edges (because the wingspan of current clips fail to close such defects), and in patients with a leak occurring in the setting of a malignant lesion (because clips tend to tear through neoplastic tissue, failing to hold the edges of the perforation).⁸ Overall therapeutic outcome depends both on successful sealing of the wall defect and the success of subsequent self-expanding metallic stent removal or self-expanding plastic stent migration.

Recently, the concept of stent-guided regeneration and re-epithelialization of digestive perforations emerged for the completion of postoperative disunion with a fully covered stent.⁹ The authors postulated that a covered stent may act as a support to guide tissue regeneration and re-epithelialization alongside the external membrane of the covered stent by stimulating mucosal regrowth. However, this concept has not yet been proven by experimental models.

The repair of chronic esophageal perforations or recalcitrant fistulas with endoscopic therapy can be obtained by targeting the site from the inside and outside. This inside-outside approach can be a useful option to ensure both external closure of the defect by mean fibrin glue, sealants, or newly developed acellular matrix graft, as well as internal protection of the breach by using stents and the OTSC system.¹⁰

A successful outcome of endoscopic sealants for esophageal fistulas is dependent upon the size of the fistula and the absence of active infection around the site of the leak, cancer, or obstruction distal to the site of the leak. Chronic fistulas benefit from clearance of mediastinal infection with debridement and drainage before endoscopic closure.¹¹ In our opinion, brushing the fistulous tract and correct placement of the sealant application catheter inside the fistulous orifice are critical for successful closure of the fistula.¹⁰

Endoscopic insertion of strips of Surgisis (Cook Biotech, Inc), an acellular matrix derived from porcine submucosa promoting fibroblast proliferation without stimulating a foreign-body reaction, has also been shown to be successful in the closure of chronic esophageal anastomotic leaks not suitable for closure by stents and fibrin glue.¹⁰

For large esophageal defects with extravisceral collection that could be endoscopically explored, vacuum-assisted closure was recently reported.¹² As with infected abdominal wounds, this method allows regular visualization of the leak and infected cavity and promotes tissue granulation to obtain a secondary-intention closure of the fistula. The regular debridement and, therefore, control of the septic focus appears to be another major advantage over stent therapy.

In conclusion, recent advances in endoscopic closure devices have increased therapeutic options in selected patients. A multidisciplinary approach involving the surgeon, endoscopist, interventional radiologist, and intensiv-ist remains mandatory in all patients with esophageal perforations in order to reduce mortality and complications. Due to the continuous implementation of technological resources at our disposal, dedicated training programs in endoscopic closure of perforations are needed.