Abstract: Objective:To investigate the risk factors of anastomotic leakage after radical gastrectomy for gastric cancer and establish a risk prediction scoring model for postoperative anastomotic leakage.
Methods:The retrospective casecontrol study was conducted. The clinicopathological data of 1 757 patients who underwent radical gastrectomy for gastric cancer in the First People′s Hospital of Yibin from February 2000 to December 2017 were collected. There were 1 207 males and 550 females, aged (59±11)years, with a range from 48 to 70 years. Observation indicators: (1) followup situations; (2) risk factors analysis of anastomotic leakage after radical gastrectomy for gastric cancer; (3) establishment of risk prediction scoring model and verification; (4) risk prediction scores and probability of anastomotic leakage in patients with different scores. Followup using outpatient examination and telephone interview to detect anastomotic leakage after radical gastrectomy and reoperation or death caused by anastomotic leakage up to June 2018. Measurement data with normal distribution were represented as Mean±SD. Count data were described as absolute number. The univariate analysis and multivariate analysis were performed using the chisquare test and Logistic regression model respectively. The discrimination and fitting degree of the model were detected by the Hosmer-Lemeshow test and receiver operating characteristic(ROC) curve.
Results: (1) Followup situations: 800 of 1 757 patients were followed up at 1 week, 1 month, 3 months and 6 months after discharge. During the followup, 75 had anastomotic leakage, 60 of which recovered after conservative treatment, 9 recovered after reoperation, 6 died of complications including septic shock and cardiac and respiratory failure. (2) Risk factors analysis of anastomotic leakage after radical gastrectomy for gastric cancer: results of univariate analysis showed that patients with diabetes, preoperative pulmonary insufficiency, preoperative level of albumin and volume of intraoperative blood loss were related factors affecting anastomotic leakage after radical gastrectomy for gastric cancer (x²=5.604, 4.975, 18.563, 35.688, P<0.05). Results of multivariate analysis showed that patients with diabetes, preoperative pulmonary insufficiency, preoperative level of albumin <30 g/L and volume of intraoperative blood loss >400 mL were independent risk factors affecting anastomotic leakage after radical gastrectomy for gastric cancer (odds ratio=2.337, 1.946, 3.478, 4.357, 95% confidence interval: 1.136-4.804, 1.022-3.705, 1.871-6.464, 2.678-7.090, P<0.05). (3) Establishment of risk prediction scoring model and verification. Risk prediction equation was established according to the multivariate Logistic regression results: P=1/1+exp(4.092-0.666*X1-0.849*X2-1.246*X3-1.472*X4). The fitting degree of the model was detected by the Hosmer-Lemeshow test (P=0.287). The discrimination of the model was detected by the ROC curve, with the area under curve as 0.734 (95% confidence interval: 0.689-0.834, P=0.002) . (4) Risk prediction scores and probability of anastomotic leakage in patients with different scores: the risk prediction scores of anastomotic leakage after radical gastrectomy for gastric cancer were 1, 1, 2, 2 in patients with diabetes,preoperative pulmonary insufficiency, preoperative level of albumin <30 g/L and volume of intraoperative blood loss >400 mL, respectively. The incidence of anastomotic leakage of patients with risk prediction scores of 0, 1, 2, 3, 4, 5, 6 was 1.6%, 3.2%, 5.9%, 10.1%, 19.3%, 31.8% and 47.6%, respectively. The incidence of anastomotic leakage was 13.7% of patients with score ≥3 and 3.5% of patients with score <3.
Conclusions:Patients with diabetes, preoperative pulmonary insufficiency, preoperative level of albumin <30 g/L and volume of intraoperative blood loss >400 mL are independent risk factors affecting anastomotic leakage after radical gastrectomy for gastric cancer. Establishment of a risk prediction scoring model for anastomotic leakage after radical gastrectomy for gastric cancer can effectively identify highrisk patients with anastomotic leakage after radical gastrectomy.