An Approach to Accelerate Healing and Shorten the Hospital Stay of Patients With Anastomotic Leakage After Esophagectomy: An Explorative Study of Systematic Endoscopic Intervention

OBJECTIVE: To explore the comprehensive role of systemic endoscopic intervention in healing esophageal anastomotic leak. METHODS: In total, 3919 consecutive patients with esophageal cancer who underwent esophagectomy and immediate esophageal reconstruction were screened. In total, 203 patients (5.10%) diagnosed with anastomotic leakage were included. The participants were divided into three groups according to differences in diagnosis and treatment procedures. Ninety-four patients received conventional management, 87 patients received endoscopic diagnosis only, and the remaining 22 patients received systematic endoscopic intervention. The primary endpoint was overall healing of the leak after oncologic esophageal surgery. The secondary endpoints were the time from surgery to recovery and the occurrence of adverse events. RESULTS: 173 (85.2%; 95% CI, 80.3-90.1%) of the 203 patients were successfully healed, with a mean healing time of 66.04 ± 3.59 days (median: 51 days; range: 13-368 days), and the overall healing rates differed significantly among the three groups according to the stratified log-rank test (P<0.001). The median healing time of leakage was 37 days (95% CI: 33.32-40.68 days) in the endoscopic intervention group, 51 days (95% CI: 44.86-57.14 days) in the endoscopic diagnostic group, and 67 days (95% CI: 56.27-77.73 days) in the conventional group. The overall survival rate was 78.7% (95% CI: 70.3 to 87.2%) in the conventional management group, 89.7% (95% CI: 83.1 to 96.2%) in the endoscopic diagnostic group and 95.5% (95% CI: 86.0 to 100%) in the systematic endoscopic intervention group. Landmark analysis indicated that the speed of wound healing in the endoscopic intervention group was 2-4 times faster at any period than that in the conservative group. There were 20 (21.28%) deaths among the 94 patients in the conventional group, 9 (10.34%) deaths among the 87 patients in the endoscopic diagnostic group and 1 (4.55%) death among the 22 patients in the endoscopic intervention group; this difference was statistically significant (Fisher exact test, P < 0.05). CONCLUSION: Tailored endoscopic treatment for postoperative esophageal anastomotic leakage based on endoscopic diagnosis is feasible and effective. Systematic endoscopic intervention shortened the treatment period and reduced mortality and should therefore be considered in the management of this disease.