Steven S Cheung, MD, Brian G Biggerstaff, MD, Pulkesh Bhatia, MBBS, Carrie Bertolloti, APRN, Pradeep K Pallati, MBBS, Kalyana C Nandipati, MBBS, Sumeet K Mittal, MBBS. Creighton University
Background:
Anastomotic leaks significantly increase morbidity and mortality following esophageal resections. Endoscopic techniques continue to evolve as viable alternatives to operative intervention. This study compares features and outcomes between conservative, endoscopic, and operative management of anastomotic leaks.
Methods:
After Institutional board review approval retrospective review of a prospectively maintained database was done to identify patients who had anastomotic leak after an esophageal resection. The charts were reviewed and data pertaining to demographics, type of procedure (open, minimally invasive, or hybrid), procedural indication, leak location (cervical, thoracic, or abdominal), management style (conservative, endoscopic, or operative), leak characteristics, and outcomes (complications, reoperations, mortality, ICU duration, and length of stay) extracted.
Results:
The anastomotic leak rate is 9.4% during the study period with direct mortality rate of 6.9% (n=2). The leak rate for cervical, thoracic, and abdominal anastomosis were 8.4%, 13.2% and 5.6% respectively.
Conservative management was utilized for 13 patients, most commonly for cervical leaks. In almost all cases, the leak was identified on initial swallow study prior to initiation of oral diet. There was one death in this cohort due to gastric conduit necrosis and resultant aortic fistula. Endoscopic management was used for 12 patients, most commonly for intra-thoracic anastomosis. Leaks were less likely to be contained and identified later compared to cervical leaks. Endoscopic techniques include stent placement (n=11) and endoscopic vacuum assisted closure (n=2). Most patients required multiple endoscopic washouts during their hospital course (58.3%). Re-operative intervention was needed in 4 patients, performed exclusively for thoracic leaks. One patient developed gastro-bronchial fistula resulting in mortality. Strictures requiring balloon dilation were most common in conservatively managed patients (46.2%,) compared to endoscopic and operative management (33.3% and 0.0% respectively). The only fistula occurred in an operatively managed patient.
ICU duration and overall length of stay were significantly shorter for conservatively managed patients compared to endoscopic or operative management (p<0.05). No statistical significance was not reached between endoscopic and operative management in our cohort.
Conclusions:
Endoscopic techniques are an effective alternative to operative intervention in management of complicated anastomotic leaks following esophagectomy. While this strategy may require intensive repeat periodic endoscopic washouts, it avoids morbidity of major re-operation.