Abstract
BACKGROUND: Minimally invasive McKeown esophagectomy is an important surgical approach for esophageal cancer. Anastomotic leak is one of its common and serious complications. We assumed that the preoperative risk factors and postoperative indicators would predict or detect anastomotic leak. METHODS: Between December 2016 and July 2017, patients underwent minimally invasive McKeown esophagectomy were identified and their preoperative variables and postoperative test indicators were recorded. Fisher's exact test, 2-tailed unpaired t test, nonparametric test and logistic regression were used to compare these datum between patients with or without anastomotic leak (AL). Receiver Operator Characteristic (ROC) curve was used to identify the best cut-off value of drainage amylase concentration for distinguishing anastomotic leak. RESULTS: In all the 96 patients included, 12 patients were diagnosed as anastomotic leak by the esophagram. No differences in preoperative variables were observed between patients with and without AL. Patients in AL group appeared to have a lower prealbumin concentration in AL group on POD (postoperative day) 4(P = 0.05), POD 5(P = 0.04), POD 6 (P = 0.06). Prealbumin concentration cutoff value of 128 g/L on postoperative day 5 is 100.00% sensitive and 50.00% specific for predicting esophageal leaks. Drain amylases levels were higher in patients with anastomotic leak than those without anastomotic leak on POD 3(P = 0.03), POD 4(P = 0.01), POD 5(P < 0.001), POD 6(P < 0.001). The drain amylase cutoff value of 85 IU/L on postoperative day 4 was 75.00% sensitive and 84.00% specific for detecting esophageal leaks; the cutoff value of 65 IU/L on postoperative day 5 was 91.67% sensitive and 80.77% specific. The cutoff of 55/L on POD 6 is 100% sensitive and 86.96% specific. CONCLUSION: Drainage amylase concentration on postoperative days may help to discover anastomotic leak in early stage after minimally invasive McKeown esophagectomy. Prealbumin concentration below 128 g/L on POD 5 might be potential risk factor for anastomotic leak.