Anatomic patterns of anastomotic leaks after Ivor Lewis esophagectomy for cancer: Impact on management and outcomes

Background: Anastomotic leakage presentation after Ivor Lewis esophagectomy may vary on imaging. Such variations may influence anastomotic leakage management and outcomes.

Methods: All consecutive patients who underwent Ivor Lewis esophagectomy for cancer between 2012 and 2019 in 2 referral centers were included. Anatomical patterns of anastomotic leakage were defined on imaging as follows: eso-mediastinal anastomotic leakage was a leak contained in the posterior mediastinum, eso-pleural anastomotic leakage was a leak involving the pleural cavity, and eso-bronchial anastomotic leakage was a leak communicating with the tracheobronchial tract. According to the Esophageal Complications Consensus Group definition, management and 90-day mortality were evaluated according to these patterns.

Results: Among 731 patients, 111 (15%) developed anastomotic leakage consisting of eso-mediastinal anastomotic leakage (n = 87, 79%), eso-pleural anastomotic leakage (n = 16, 14%) and eso-bronchial anastomotic leakage (n = 8, 7%). There was no difference among these groups regarding preoperative characteristics or time to anastomotic leakage diagnosis. There was a significant difference in initial management according to anastomotic leakage anatomic patterns (P = .001). More than half of patients who experienced eso-mediastinal anastomotic leakage (n = 46, 53%) were initially treated conservatively without requiring intervention (Esophageal Complications Consensus Group type I), whereas most patients with eso-pleural anastomotic leakage (n = 14, 87.5%) and all with eso-bronchial anastomotic leakage (n = 8, 100%) initially required interventional or surgical treatment (Esophageal Complications Consensus Group type II-III). Anastomotic leakage anatomic patterns had a statistically significant impact on 90-day mortality, intensive care unit stay, and total hospital stay (P < .001).

Conclusion: Anastomotic leakage anatomic patterns after Ivor Lewis esophagectomy influence outcomes. Further studies are warranted to validate it in a prospective setting. Anastomotic leakage anatomic patterns may help in guiding anastomotic leakage management.