Randomized clinical trial of ultrasonic dissector or conventional division in distal pancreatectomy for non-fibrotic pancreas

Abstract

Background: Resection of the non-fibrotic pancreas is prone to postoperative pancreatic fistula because of well preserved exocrine secretions and easily crushed soft parenchyma. The purpose of this study was to evaluate ultrasonic dissection for division of the non-fibrotic pancreas in distal pancreatectomy. Methods: All pancreata included in this study were soft on direct palpation and their main ducts had no dilatation, at least proximally from the transection line. Fifty-eight patients with gastric cancer or pancreatic disease were randomly assigned to the two groups. In the ultrasonic dissection (UD) group (n = 27), all pancreatic ducts were identified and ligated securely. The stump was left open without parenchymal suturing. In the conventional (CV) group (n = 31), the pancreas was cut with a knife and the stump was oversewn in mattress fashion. The main pancreatic duct was ligated in all patients in both groups. Pancreatic fistula was defined as a pancreatic fluid discharge for more than 7 days after operation diagnosed according to amylase concentration in the drainage fluid. Results: In the UD group, approximately 20–30 tubes including a mean(s.d.) 5·2(0·8) (range 4–6) pancreatic ducts were skeletonized and ligated per patient. There were nine pancreatic fistulas (16 per cent); one in the UD group and eight in the CV group (P= 0·020). Conclusion: In distal pancreatectomy for the non-fibrotic pancreas, ultrasonic dissection without suture closure of the stump reduced the incidence of pancreatic fistula compared with conventional division and suture, in this randomized trial.