Management of anastomotic leakage is challenging for patients received esophagectomy The mortality is high, however, the optimal treatment still need to be determined [2, 4, 11]. Various conservative treatment protocols have been reported for the treatment of anastomotic leakage over the past two decades, including the application of biodegradable fistulae plugs or fibrin glue, endoscopic transluminal drainage or clipping and metallic esophageal stent insertion [4, 8,9,10]. Surgical repair of the anastomotic leakage is the traditional protocol, such as, esophagectomy or thoracotomy and repair . Despite of the advances in surgical technique, the overall mortality remains as high as 20 to 50% [3,4,5,6]. Nowadays, treatment of choice in the first line is an endoscopic approach.
Esophageal stents were initially served as a palliative treatment for patients with esophageal carcinoma. Currently, esophageal stents have been used to treat esophageal benign diseases [13,14,15,16,17]. Metallic stents are usually used for treatment of tumor stenosis or esophageal perforation [18, 19]. Successful and effective management of anastomotic leakage needs prompt elimination of contamination by covered stent placement, and adequate drainage of the abscess cavity. We present 24 consecutive patients treated with three-tube method and covered stent placement for anastomotic leakage. Our clinical outcomes indicated that this interventional method can easily be performed under fluoroscopic guidance. No perioperative death was observed, which is lower than previous reports [6, 15, 16, 20]. After covered stent placement, the leakage is still allowed to continuously drain fluid via abscess drainage tube. In our study, all patients received continue abscess drain for a median duration of 2.6 months. Drainage of an abscess cavity is also possible percutaneously under CT scan control and it is generally easier for the subsequent follow-up. In this study, three patients were in need of external thoracic drainage. Compared with endoscopic drainage, radiologic drainage can be performed via the leaks without the need of percutaneous puncture.
The duration between esophageal surgery and leakage diagnosis is essential for the clinical outcomes . The median interval between surgery and leakage was 0.4 months. The clinical outcomes were favorable; abscess cavity was markedly decreased in 8 patients and disappeared in 16 cases. Treatment results may be related to the size of the leak, and effective results of biodegradable fistulae plugs or fibrin glue application are generally observed only in small leaks.
Certain complications can be found in our interventional protocol. Stent migration is a common complication, especially in patients without esophageal stricture [21, 22]. All esophageal stents used in our study were coved ones, which may account for high rate of stent migration. Nine patients showed stent migration and were adjusted or replaced for 1 to 3 times. Only 1 patient showed migration of drainage tube, however, drainage tubes were regularly adjusted and replaced for 0 to 6 times to achieve effective drainage during follow up. The abscess drainage tubes were adjusted or exchanged for a median time of 2 times. Besides, esophageal stents and abscess drainage tubes were successfully removed from 14 patients, without difficulties of removal or severe complications. Recovery lines in the proximal end of stent are used for stent fixation to avoid stent migration, and for the adjustment or recovery of the migrated stent. Appropriate size of stent should been used, considering that small size of stent is prone to migrate. Of course, clips and stent with flaps can be used to reduce the migration rate.
There were some limitations. This was a retrospective study with relatively small number of enrolled patients. The esophageal stents and abscess drainage tubes needed adjusted or replaced repeatedly during follow up. BMI data had not collected previously considering that BMI may be not closely related to our treatment. We had not measured how much negative pressure and the study interval was long.