[Clinical features and risk factors of anastomotic leakage after radical esophagectomy].

OBJECTIVE: To analyze the clinical features and risk factors of anastomotic leakage after radical esophagectomy of esophageal carcinoma. METHODS: The clinical data of 547 esophageal cancer patients underwent radical esophagectomy in Tianjin Medical University Cancer Hospital from January 2012 to December 2013 was analyzed retrospectively. There were 421 male and 126 female patients, with a median age of 65 years (ranging from 29 to 82 years). There were 155 cases of upper esophageal carcinoma, 340 cases of middle esophageal carcinoma and 52 cases of lower esophageal carcinoma. The surgical procedures included 41 cases completed through Sweet, 145 cases completed through McKeown, 279 cases completed through Ivor Lewis, 82 cases completed through minimally invasive esophagectomy. Moreover, 24 of 547 cases underwent preoperative neoadjuvant radiochemotherapy. χ² test and Cox's proportional hazards regression model were used for univariate analysis and multivariate analysis of the risk factors of postoperative anastomotic leakage. RESULTS: Twenty-seven of 547 cases with esophagectomy occurred anastomotic leakage and the incidence rate was 4.94% (27/547). One of 27 cases died and the mortality rate was 3.70% (1/27). The time of anastomotic leakage found was 4 to 45 days, with a median time of 10 days. There were 0 case of early leakage, 20 cases of mid-term leakage, 7 cases of late leakage. Three of 27 cases with anastomotic leakage had tracheoesophageal fistula, while 3 cases had contralateral pleural fistula. As to the incidence rate of anastomotic leakage, there was statistically significant difference between cervical anastomotic leakage (8.14%, 18/221) and intrathoracic anastomotic leakage (2.76%, 9/326) (χ² =7.41, P=0.000), among Sweet (4.88%, 2/41), McKeown (9.66%, 14/145), Ivor Lewis (2.51%, 7/279) and MIE (4.88%, 4/82) (χ² =21.48, P=0.000), and between with (16.67%, 4/24) and without (4.40%, 23/523) neoadjuvant radiochemotherapy (χ² =9.20, P=0.000). The multivariate analysis showed that anastomotic site (HR=2.594, P=0.048), surgical approach (HR=5.689, P=0.003) and preoperative neoadjuvant radiochemotherapy (HR=3.604, P=0.027) are independent risk factors for anastomotic leakage after esophagectomy. CONCLUSIONS: The mid-term anastomotic leakage after esophagectomy occurs higher. McKeown is a main surgical procedure and neoadjuvant radiochemotherapy is an important factor for the anastomotic leakage.