Endoscopic management of esophageal anastomotic leaks and perforations with the use of esophageal stents is technically feasible. It seems to be safe and effective when performed along with mediastinal or pleural drainage. Esophageal stent can, therefore, be considered as a treatment option in the management of patients who present early after esophageal perforation or anastomotic leak with limited mediastinal or pleural contamination.

Aims This updated systematic review aimed to assess the safety and effectiveness of oesophageal stents in these two scenarios. Background Endoscopic placement of oesophageal stents may be used in benign oesophageal perforation and oesophageal anastomotic leakage to control sepsis and reduce mortality and morbidity by avoiding thoracotomy. Methods A systematic literature search of all published studies reporting use of metallic and plastic stents in the management of postoperative anastomotic leaks, spontaneous and iatrogenic oesophageal perforations were identified. Primary outcomes were technical (deploying ≥1 stent to occlude site of leakage with no evidence of leakage of contrast within 24-48 hours) and clinical success (complete healing of perforation or leakage by placement of single or multiple stents irrespective of whether the stent was left in situ or was removed). Secondary outcomes were stent migration, perforation and erosion and mortality rates. Subgroup analysis were performed for plastic versus metallic stents and anastomotic leaks versus perforations separately. Results A total of 66 studies (n=1752 patients) were included. Technical and clinical success rates were 96% and 87%, respectively. Plastic stents had significantly higher migration rates (24% vs 16%, p=0.001) and repositioning (11% vs 3%, p<0.001) and lower technical success (91% vs 95%, p=0.032) than metallic stents. In patients with anastomotic leaks, plastic stents were associated with higher stent migration (26% vs 15%, p=0.034), perforation (2% vs 0%, p=0.013), repositioning (10% vs 0%, p<0.001) and lower technical success (95% vs 100%, p=0.002). In patients with perforations only, plastic stents were associated with significantly lower technical success (85% vs 99%, p<0.001). Conclusions Covered metallic oesophageal stents appear to be more effective than plastic stents in the management of oesophageal perforation and anastomotic leakage. However, quality of evidence of generally poor and high-quality randomised trial is needed to further evaluate best management option for oesophageal perforation and anastomotic leakage.