[Overlap esophagojejunostomy with multi-mode modifications in totally laparoscopic total gastrectomy: safety and feasibility of 152 cases from a single center].

Currently, the Overlap anastomosis is one of the most favored reconstruction methods of intracorporeal esophagojejunostomy (EJS). Despite many advantages of the method, it remains some shortcomings to be improved when it comes to the retraction of the esophagus stump, the insertion of the anvil fork of the linear stapler into a "pseudo" lumen, and the closure of the common entry hole. This study aims to investigate the safety and feasibility of a multi-mode modified Overlap anastomosis. A descriptive case series study was conducted. Medical records of 152 consecutive patients who underwent totally laparoscopic total gastrectomy (TLTG) with our multi-mode modified Overlap EJS method by the same surgical team at our department from February 2017 to June 2020 were retrospectively analyzed. The multi-mode modified Overlap method mainly included (1) After ensuring the safety of tumor resection margin (proximal margin was at least 3 cm from the tumor), the esophagus was partially transected from left to right (with 5-8 mm width esophagus continuation). The specimen was then placed in a plastic bag which was tied up at the mouth using strings with a part of the esophageal wall poking through. Then the plastic bag containing the specimen was transferred to the right lumbar region, while the patient's body position was adjusted so that the abdominal esophagus could be pulled by the gravity of the specimen. (2) Using the "three-direction traction" method. The esophageal lumen was properly exposed, then guided by the gastric tube, the anvil fork was accurately placed into the esophageal lumen for completing the side-to-side EJS. (3) The 3-0 barbed suture was used in the closure of the common entry hole of the stapler from dorsally to ventrally with simple one-layer continuous suture (the stitch going from inside to inside) followed by continuous Lembert's suture (the stitch going from outside to outside). Combined with clinicopathological characteristics, the perioperative outcomes and postoperative complications of the whole group were analyzed and evaluated. The study cohort included 129 men and 23 women, with a mean age of (60.2±9.1) years and a mean body mass index (BMI) of (23.2±3.1) kg/m(2). Of the 152 patients, 23 patients (15.1%) had a history of previous abdominal surgery; dentate line was invaded by tumor in 21 patients (13.8%). The mean length of the proximal resection margin was (3.3±0.3) cm and the postoperative pathological examination indicated negative resection margin tumor. The mean operative time and anastomotic time were (302.1±39.9) minutes and (29.8±5.4) minutes, respectively. The mean estimated blood loss was (87.9±46.4) ml. The mean length of postoperative hospital stay was (12.3±7.3) days. The overall severe postoperative complications (Clavien-Dindo ≥ II) occurred in 22 patients (14.5%). Six cases of pancreatic leakage were successfully recovered by adequate drainage, inhibition of pancreatic exocrine secretion and nutritional support. Ten cases of pneumonia and three cases of abdominal infection were cured with anti-infection and physical therapy. Two patients developed anastomotic leakage postoperatively. One case was caused by excessive tension of the Roux loop of the jejunum and excessive opening on the side of the jejunum after side-to-side anastomosis, and the other case was caused by an accidental intraoperative occurrence of "nasogastric tube stapled to the side-to-side anastomosis". Both of them recovered after conservative treatment including adequate drainage, anti-infection, and adequate nutritional support. One patient underwent immediate open surgery because of Peterson's hernia 7 days after TLTG, and the patient died due to extensive small bowel necrosis. Multi-mode modified overlap method simplifies the operation and reduces the difficulty of EJS. It is a safe and feasible method for EJS.