Adverse events of upper GI endoscopy

The most common adverse events of esophageal dilation are perforation and bleeding. Wire-guided bougie dilation or through-the-scope balloon dilation may have lower risks of adverse events than blind passage of dilators. Randomized trials suggest that wire-guided polyvinyl dilators and through-the-scope balloons have similar rates of both efficacy and adverse events.^{,  ,  ,} 

The rate of perforation after esophageal dilation for esophageal rings and simple peptic strictures is lower than that of certain high-risk lesions. Dilation of complex strictures (angulated, multiple, or long) with Maloney dilators may be associated with a 2% to 10% risk of perforation^,  so wire-guided or balloon dilation is likely a safer alternative. Dilation of caustic strictures, which tend to be long and angulated, is associated with a higher rate of adverse events.^,  Dilation of eosinophilic esophagitis is associated with a high incidence of mucosal tears, but only 1 perforation was identified in a systematic review of 671 dilations for eosinophilic esophagitis. The risk of perforation resulting from dilation of malignant strictures of the esophagus is approximately 10%^,  and is associated with increasing dilator diameter.^{,  ,  ,}  Radiation-induced strictures have also been reported to have a high rate of dilation-related adverse events, but this risk may be related to the presence of malignancy rather than the effect of radiation.

Pain is the most common symptom related to perforation.^{,  ,  ,}  Fever, crepitus, pleuritic chest pain, leukocytosis, and pleural effusion may also be present. Perforation with associated air dissection may be diagnosed by plain radiography of the neck and/or chest, but such findings may be absent immediately after perforation. If a perforation is suspected, contrast esophagography should be performed, usually beginning with water-soluble contrast. If the site of perforation cannot be determined but suspicion remains high, a barium esophagram or CT scan of the chest is indicated. A CT scan with oral contrast is sensitive for the site of perforation and for more subtle findings such as minute amounts of air or fluid.

The approach to the patient with perforation depends on the state of health of the individual, the site of the perforation, and the overall prognosis. In selected patients, early recognition may allow nonoperative management with nasogastric suction, intravenous antibiotics, and parenteral nutrition. Surgical consultation should be obtained, and surgical management is recommended for larger perforations in which the pleural space is involved or for failure to respond to medical management.^,  Case series of successful endoscopic closure of esophageal perforation with endoluminal stents, endoscopic clips, or suturing devices have been published,^{,  ,  ,  ,}  although comparative data are lacking.

Pneumatic dilation of the lower esophageal sphincter is associated with increased risk of postprocedure pain, aspiration, bleeding, and perforation.^,  The rate of perforation is between 1.6% and 8%.^,  The risk of perforation may be lower when interval, graded dilation is used, beginning with a 30-mm diameter balloon and progressing to larger diameter balloons, only if symptoms do not improve. Using this technique, the overall risk of perforation is reported to be less than 2%.^{,  ,}  Contrast esophagography should be performed for patients with persistent postprocedure pain, tachycardia, fever, or subcutaneous crepitus. Nonoperative management with nasogastric tube decompression and intravenous antibiotics may be used for contained perforations caused by pneumatic dilation. Perforations resulting in extravasation of contrast during postprocedure esophagography may require operative intervention.

Endoscopic balloon dilation for benign gastric outlet obstruction has been associated with perforation rates as high as 7.4%.^{,  ,  ,  ,  ,}  Risk factors for perforation include dilation in the setting of active ulceration and dilation with balloons greater than 15 mm in diameter.^{,  ,  ,}  Graded dilation with stepwise increase of balloon size has been suggested to help reduce the risk of perforation.^, 

Endoscopic polypectomy in the UGI tract is associated with low rates of pain, bleeding, and perforation. Immediate bleeding after gastric polypectomy is more common than bleeding after polypectomy at other sites, with rates ranging from 3.4% to 7.2%.^{,  ,  ,}  Delayed bleeding after polypectomy of duodenal adenomas is reported in 3.1% to 22% of patients.^{,  ,} 

EMR is used to excise focal lesions of the mucosa and involves resection into the submucosal layer. Common self-limited adverse events of EMR include chest pain, abdominal pain, dysphagia, odynophagia, and dyspepsia. The overall incidence of serious adverse events such as bleeding, perforation, and stricture has been estimated to be between 0.5% and 5%. Bleeding occurs more often with multifocal EMR and with EMR of gastric lesions.^{,  ,}  Perforation with gastric EMR is reported more frequently than with esophageal EMR, possibly because of the larger lesions encountered in the stomach. Stricture formation is mostly reported after esophageal EMR, especially when circumferential resection is performed. The incidence of esophageal stricture after focal EMR is less than 0.5%, compared with an incidence of 12% to 35% when more than 50% of the esophageal circumference is resected.^, 

Endoscopic submucosal dissection (ESD) allows for en bloc excision of large mucosal lesions of the GI tract by using a variety of specialized accessories.^,  Adverse events of ESD are similar to those of EMR, but occur with greater frequency given the larger areas of resection. The overall incidence of bleeding and perforation with ESD is 11% and 6%, respectively.^{,  ,  ,  ,}  Asymptomatic pneumomediastinum may occur in as many as 31% of ESDs and is of uncertain clinical significance.

Ablation of mucosal lesions of the UGI tract can be performed with a variety of devices including heater probes, multipolar electrocoagulation, argon plasma coagulation (APC), and Nd-YAG laser. Self-limited adverse events commonly reported include pain, dysphagia, and nausea. The incidence of serious adverse events associated with APC appears to be higher than that of other modalities, especially when treating long segments of Barrett's esophagus or with multiple sessions of ablation.^{,  ,  ,}  Randomized trials with APC report bleeding rates of as high as 4%, esophageal perforation in as many as 2% of patients, and stricture formation in as many as 6% of patients.^{,  ,} 

Photodynamic therapy (PDT) with porfimer sodium as a photosensitizing agent is used for palliation of dysphagia in advanced esophageal cancer and for ablation of Barrett's epithelium with high-grade dysplasia. PDT of the esophagus frequently causes chest pain, fever, and pleural effusion.^,  PDT with porfimer sodium results in esophageal stricture formation in 11% to 42% of patients.^,  Photosensitivity reactions occur in 10% to 60% of patients.^, 

Radiofrequency ablation (RFA) of Barrett's epithelium has a relatively favorable adverse event profile. In 1 randomized trial, the degree of chest discomfort was higher after RFA than in the control group, but resolved within 8 days of the procedure. Superficial lacerations have been noted during 6% of procedures, but bleeding requiring endoscopic therapy occurred in less than 2% of procedures.^{,  ,  ,}  The incidence of RFA-associated esophageal stricture ranges from 2% to 8%.^{,  ,}  Procedure-related perforation has been reported.

Cryotherapy has not been as well studied to date. Small case series report common self-limited symptoms such as pain and dysphagia. The incidence of strictures ranges between 4% and 10%.^{,  ,}  Esophageal perforation was reported in 1 patient with Marfan syndrome undergoing liquid nitrogen cryotherapy.

Stents may be deployed endoscopically to achieve luminal patency in any part of the UGI tract. Rigid esophageal stents are no longer used and have been replaced by self-expanding stents.^,  Immediate adverse events of esophageal self-expandable metal stents (SEMSs) occur in 2% to 12% of patients and include aspiration, respiratory compromise caused by tracheal compression, improper positioning, and perforation.^{,  ,  ,}  Immediate adverse events may be minimized by adequate patient preparation and positioning, familiarity of the endoscopist with the stent mechanism and characteristics, the use of soft-tipped guidewires, and avoidance of aggressive prestent dilation.^,  Early postdeployment adverse events, such as chest pain and nausea, are common and resolve with conservative measures in most cases.^{,  ,}  Significant bleeding after SEMS placement is not common, but may be life-threatening. Late adverse events after esophageal SEMS placement occur in 20% to 40%. Pyrosis and regurgitation are common when the gastroesophageal junction is bridged with a stent. Strict antireflux measures, high-dose acid suppression, and the use of stents designed to prevent reflux have been used with varying degrees of success.^{,  ,}  Recurrent occlusion of SEMS is reported in as many as 30% of patients and can occur because of tumor overgrowth, tissue hyperplasia at the ends of the stent, stent migration, or food impaction. The use of covered stents reduces the risk of tumor ingrowth.^,  Occlusion by tissue may be treated by endoscopic ablation of the tissue or placement of a second stent. Food impactions may be managed endoscopically. Late perforation of the esophagus caused by ischemia of the esophageal wall and tracheoesophageal fistulae have been reported.^{,  ,}  Pretreatment with chemoradiotherapy has been reported to increase the incidence of adverse events of esophageal SEMSs by some authors but not by others.^,