Effect of duodenal stump reinforcement on postoperative complications in patients undergoing laparoscopic radical gastrectomy

Abstract

Objective: To evaluate the influence of duodenal stump reinforcing on the short-term complications after laparoscopic radical gastrectomy. Methods: A retrospective cohort study with propensity score matching (PSM) was conducted. Clinical data of 1204 patients with gastric cancer who underwent laparoscopic radical gastrectomy at the First Affiliated Hospital of Army Medical University from April 2009 to December 2018 were collected. The digestive tract reconstruction methods included Billroth II anastomosis, Roux-en-Y anastomosis and un-cut-Roux- en-Y anastomosis. A linear stapler was used to transected the stomach and the duodenum. Among 1204 patients, 838 were males and 366 were females with mean age of (57.0±16.0) years. Duodenal stump was reinforced in 792 cases (reinforcement group) and unreinforced in 412 cases (non-reinforcement group). There were significant differences in resection range and anastomotic methods between the two groups (both P<0.001). The two groups were matched by propensity score according to the ratio of 1∶1, and the reinforcement group was further divided into purse string group and non-purse string group. The primary outcome was short-term postoperative complications (within one month after operation). Complications with Clavien-Dindo grade ≥ III a were defined as severe complications, and the morbidity of complication between the reinforcement group and the non-reinforcement group, as well as between the purse string group and the non-purse string group was compared. Results: After PSM, 411 pairs were included in the reinforcement group and the non-reinforcement group, and there were no significant differences in baseline data between the two groups (all P>0.05). No perioperative death occurred in any patient.The short-term morbidity of postoperative complication was 7.4% (61/822), including 14 cases of anastomotic leakage (23.0%), 11 cases of abdominal hemorrhage (18.0%), 8 cases of duodenal stump leakage (13.1%), 2 cases of incision dehiscence (3.3%), 6 cases of incision infection (9.8%) and 20 cases of abdominal infection (32.8%). Short-term postoperative complications were found in 25 patients (6.1%) and 36 patients (8.8%) in the reinforcement group and the non-reinforcement group, respectively, without significant difference (χ2=2.142, P=0.143). Nineteen patients (2.3%) developed short-term severe complications (Clavien-Dindo grade ≥IIIa), while no significant difference in severe complications was found between the two groups (1.7% vs. 2.9%, χ2=1.347, P=0.246). Sub-group analysis showed that the morbidity of short-term postoperative complication of the purse string group was 2.6% (9/345), which was lower than 24.2% (16/66) of the non-purse string group (χ2=45.388, P<0.001). Conclusion: Conventional reinforcement of duodenal stump does not significantly reduce the incidence of duodenal stump leakage, so it is necessary to choose whether to reinforce the duodenal stump individually, and purse string suture should be the first choice when decided to reinforce.