ABSTRACT
BACKGROUND AND AIMS: Endoscopic stenting is the standard of care for full thickness esophageal wall defects. The aim of this study is to evaluate outcomes of endoscopic closure of esophageal defects using stenting, with or without endoscopic suturing. METHODS: This is a single-center retrospective study of patients with esophageal wall defects who underwent endoscopic interventions. Outcomes of stenting with or without endoscopic suturing of the defect were assessed. Univariate and multivariate logistic regression models were used to examine factors associated with successful defect closure. RESULTS: One hundred and fourteen patients with esophageal wall defects underwent 254 endoscopies with an overall complete closure rate of 75.8%. Twenty-three (20.2%) patients underwent primary closure using endoscopic suturing and subsequent esophageal stenting, while 91 (79.8%) underwent esophageal stenting only. The dual modality group (versus the stent-only group) had similar defect closure rates (84.2 vs. 73.8%, p = 0.55) and time to stent migration (37 vs. 12.5 days, p = 0.07), but was associated with longer procedure times (60 vs. 36 min, p < 0.01) and fewer additional endoscopic procedures (13.6 vs. 43.2%, p = 0.01). Stent suturing significantly decreased migration (35.5 vs. 58.5%, p = 0.04), was associated with fewer additional endoscopies (15.4 vs. 50%, p < 0.01) and reduced need for additional stents (7.7 vs. 34.3%, p < 0.01). On multivariate analysis, chronic defects (> four weeks old) were 81% less likely to close compared to acute (≤ 4 weeks) defects (OR 0.19, CI 0.04-0.77, p = 0.02), and large diameter stents (23 mm) were associated with higher odds of defect closure (OR 3.36, CI 1.02-11.4, p = 0.04). CONCLUSION: Endoscopic treatment of esophageal wall defects is safe, effective, and more likely to be successful in acute defects using larger caliber stents. Stent suturing reduces migration, need for additional endoscopic procedures, and stent exchanges. Further comparative studies with larger cohorts are needed to validate our results.
