Pressure Necrosis for the Removal of Embedded Self Expanding Esophageal Stents: A Rare Treatment for a Common Problem

Abstract

Purpose: Covered self expandable metallic stents (SEMS) have traditionally been used for the treatment of malignant esophageal obstruction; however, recently they have been employed in the treatment of esophageal fistulas. Partially covered SEMS have been shown to be safe and effective and offer the benefit of decreased migration. However, safe removal is difficult due to tissue hyperplasia at the uncovered ends of the stent leading to its embedment into the esophageal wall. We report a case demonstrating successful removal of a partially covered SEMS by pressure necrosis created from placing a fully covered polyflex stent via a “stent in stent” method. A 64-year-old male was referred to our service after development of an esophageal perforation resulting in pneumonia and pleural effusion secondary to a TEE performed during cardioversion for paroxysmal atrial fibrillation. This arrhythmia occurred after the patient had a gastric sleeve procedure for morbid obesity. The patient had a partially covered SEMS placed with the proximal end above the gastroesophageal junction at 39cm from the incisors. The partially covered stent was initially chosen to help create a tissue seal at both ends of the stent to promote healing and to prevent migration. Tissue overgrowth was seen at the proximal end of the stent, but no tissue hyperplasia was seen at the distal end. The distal end of the stent extended down into the proximal gastric body, and the area of the perforation still appeared to be well covered. A 12 cm long fully covered polyflex stent with a central diameter of 18mm flaring out to 23mm at the ends was then successfully deployed under fluoroscopic guidance inside of the previously placed stent with the proximal end overlapping the proximal end of the previously placed stent by 3cm. We then confirmed endoscopically that the polyflex stent was completely covering the area of tissue overgrowth at the proximal end of the stent. Three weeks later, the patient went for repeat EGD, and the polyflex stent was successfully removed with alligator forceps revealing that nearly all of the tissue overgrowth had resolved at the proximal end of the stent from prior exam due to effective pressure necrosis. We initially were unable to remove the partially covered stent by gripping the proximal end with the alligator forceps, and we eventually had to grip the distal end and employ an eversion technique under fluoroscopic guidance with gentle traction to successfully remove the stent. The stent-in-stent technique is safe and effective for the removal of partially covered SEMS that are embedded into the esophageal wall. We recommend further larger prospective trials to analyze this method for the removal of SEMS.