To investigate the risk factors of cervical esophagogastric anastomotic fistula after esophagectomy of esophageal cancer.

The retrospective case-control study was conducted. The clinicopathological data of 956 patients who underwent esophagectomy and cervical esophagogastrostomy from January 2012 to December 2016 in the First Affiliated Hospital of Zhengzhou University were collected. Patients underwent Sweet or Mckeown surgery. Observation indicators: (1) intra- and post-operative situations; (2) the risk factors analysis of cervical esophagogastric anastomotic fistula after esophagectomy; (3) follow-up situations. Follow-up using outpatient examination and telephone interview was performed to detect the esophagogastric anastomotic stenosis of patients up to February 2017. Measurement data with normal distribution were represented as the ±s. Univariate analysis and comparison of count data were done using the chi-square test or Fisher exact probability method. Multivariate analysis was done using the Logistic regression model.

(1) Intra- and post-operative situations: all the 956 patients underwent successful operations, including 107 with Sweet operation and 849 with Mckeown operation. Of 956 patients, 336 received thoracotomy and 620 received thoracoscopic surgery. Tumors located in upper, middle and lower esophagus were respectively detected in 143, 627 and 186 patients. Operation time, volume of intraoperative blood loss and number of lymph node dissected in 956 patients were (274±67)minutes, (210±167)mL and 18±11, respectively. Of 956 patients, 117 had cervical esophagogastric anastomotic fistula, with an incidence of anastomotic fistula of 12.24% (117/956). Of 117 patients with cervical esophagogastric anastomotic fistula, 2 had early stage fistula, 110 had middle stage fistula and 5 had later stage fistula; 12 were cured by two-tube method (stomach tube and nutrition tube), 24 were cured by three-tube method (stomach tube, nutrition tube and chest tube or mediastinal tube), 43 were cured by open neck incision dressing, 15 were cured by fistula cavity drainage and 17 were cured by esophageal stent implantation. Sixteen patients died in hospital postoperatively, including 6 with cervical esophagogastric anastomotic fistula and 10 without cervical esophagogastric anastomotic fistula. Duration of hospital stay of 956 patients was (16±11)days, and durations of hospital stay of patients with and without cervical esophagogastric anastomotic fistula were (39±19)days and (13±6)days. Postoperative pathological examinations: 873, 9 and 74 patients were respectively diagnosed with squamous cell carcinoma, adenocarcinoma and other types of cancer. TNM staging: stage 0, Ⅰ, Ⅱ, Ⅲ, Ⅳ and unidentified stage were respectively detected in 135, 110, 325, 376, 1 and 10 patients. (2) The risk factors analysis of cervical esophagogastric anastomotic fistula after esophagectomy: univariate analysis showed that gender, age, history of diabetes, surgical method, tubular stomach production, operation time, postoperative pulmonary infection and postoperative aspirating sputum through fiberbronchoscope were risk factors affecting cervical esophagogastric anastomotic fistula after esophagectomy, with statistically significant differences (χ²=4.179, 6.174, 4.427, 4.377, 6.266, 7.057, 55.036, 51.806, P<0.05). Multivariate analysis showed that tubular stomach production, postoperative pulmonary infection and aspirating sputum through fiberbronchoscope were independent risk factors affecting cervical esophagogastric anastomotic fistula after esophagectomy, with statistically significant differences (OR=1.922, 2.907, 2.323, 95% confidence interval: 1.203-3.070, 1.682-5.023, 1.235-4.370, P<0.05). (3) Follow-up situations: 908 of 956 patients were followed up for 2-62 months, with a median follow-up time of 28 months. During the follow-up, 21 of 111 patients with cervical esophagogastric anastomotic fistula were complicated with cervical esophagogastric anastomotic stenosis, 59 of 797 patients without cervical esophagogastric anastomotic fistula were complicated with cervical esophagogastric anastomotic stenosis, showing a statistically significant difference in cervical esophagogastric anastomotic stenosis (χ²=16.803, P<0.05).

Tubular stomach production, postoperative pulmonary infection, postoperative aspirating sputum through fiberbronchoscope are independent risk factors affecting cervical esophagogastric anastomotic fistula after esophagectomy.

Esophageal neoplasms; Neck; Esophagogastric anastomotic fistula; Risk factors