Upper gastrointestinal perforations, fistula, and anastomotic leaks are severe conditions with high mortality. Temporary endoscopic placement of fully covered self-expanding metal stent (fSEMS) has emerged as treatment option. Stent migration is a major drawback of currently used stents. Migration is often attributed to a relatively too small stent diameter as esophageal stents were initially intended for the treatment of strictures. This study aimed to investigate the safety and efficacy of a large-diameter fSEMS for treatment of these conditions. Data were retrospectively collected from patients who received this stent in the Netherlands between March 2011 and August 2013. Clinical success was defined as sufficient leak closure after stent removal as confirmed by endoscopy or X-ray with oral contrast without surgical intervention or placement of another type of stent. Adverse events were graded according a standardized grading system. Stent placement was performed in 34 patients for the following indications: perforation (n = 6), anastomotic leak (n = 26), and fistula (n = 2). Technical success rate was 97% (33/34). Clinical success rate was 44% (15/34) after one stent and 50% (17/34) after an additional stent. There were no severe adverse events and stent-related mortality. The overall adverse event rate was 50% (all graded 'moderate'). There were 14 (41%) stent migrations (complete n = 8, partial n = 6). Other adverse events were bleeding (n = 2) and aspiration pneumonia (n = 1). Reinterventions for failure of the large-diameter fSEMS were placement of another type of fSEMS (n = 4), surgical repair (n = 3), or esophagectomy (n = 1). Eleven patients (32%) died in-hospital because of persisting intrathoracic sepsis (n = 10) or preexistent bowel ischemia (n = 1). This study suggests that temporary placement of a large-diameter fSEMS for the treatment of upper gastrointestinal perforations, fistula, and anastomotic leaks is safe in terms of severe adverse events and stent-related mortality. The larger diameter does not seem to prevent stent migration.

No abstract