Endoscopic vacuum therapy in the management of anastomotic insufficiency after pancreaticoduodenectomy.

A 68-year-old patient with a pT3pN2 duodenal adenocarcinoma underwent pancreaticoduodenectomy. Because the jejunal loop was anastomosed with a previous gastroenterostomy, endoscopic examination of the pancreatic and biliary anastomoses was easily feasible. On the 5th postoperative day, the patient developed septic shock and multiorgan failure; bilious secretions were noted coming from the abdominal drains. An immediate revision laparotomy was performed, at which generalized peritonitis was found. It was associated with total dehiscence of the biliojejunal anastomosis and subtotal dehiscence of the pancreaticojejunal anastomosis. Operative reconstruction of the anastomoses was impossible. Four laparotomieswith lavage of the peritoneal cavity were scheduled at 48-hour intervals. During this time, an acute hepatic artery bleed caused by septic erosion occurred. During the third laparotomy, intracavitary vacuum therapy was also started. This consists of a drainage system made by a size-adjusted, open-pore polyurethane sponge (Endo-SPONGE; B.Braun, Melsungen, Germany) that is fixed to the distal end of a gastric tube (Ventrol, 14Fr×120cm; Covidien-Nellcor, Boulder, Colorado, USA) and connected to an electronic vacuum device that generates 125-mmHg continuous negative pressure (V.A.C. Freedom; KCI, San Antonio, Texas, USA). We differentiate between intracavitary and intraluminal vacuum therapy and have described both therapeutic modalities and their clinical results in a previous publication [1]. A small percutaneous vacuum drain was placed directly onto the region of subhepatic anastomotic insufficiency. To promote compartmentalization of the septic focus, colonic mesentery was placed ventral to the vacuum sponge drainage. Biliary and pancreatic secretions were evacuatedmainly by the vacuum drainage system but also partly by additional local gravity drainage. These measures allowed us to compartmentalize the septic focus and effectively drain the biliary and pancreatic secretion to stop further intra-abdominal inflammation. The next aim was to functionally reconstruct the anastomoses. We used the principle of intraluminal vacuum therapy and endoscopically placed a vacuum drainage system via the jejunal loop that was covering the dehiscent pancreaticojejunal and biliojejunal anastomoses. Thus, it was possible to drain the digestive juices into the jejunal lumen and the additional gravity drainage showed a steady decrease in flow rate. Subsequent adjustments were made to the endoscopic vacuum drainage system solely by endoscopic means using the procedure described in our previous publications [2]. The sponge tip was placed in contact with the bile duct and pancreatic tail through the jejunal defects, simultaneously draining both (●" Fig.1). In the course of the vacuum therapy, the wound cavity developed stable granulation tissue and gradually reduced in size. It was possible to perform the six changes of the sponge by endoscopic means on the intensive care unit, without the requirement for a further laparotomy. After two periods of treatment over a total of 28 days, endoscopic vacuum therapy wasdiscontinued. Themarkers of systemic inflammation had returned to normal and pancreatic and biliary secretions were draining into the jejunal loopwithout leakFig.2 Endoscopic appearance 4 weeks later showing normalization of the anatomic situation. Fig.1 Endoscopic view showing the sponge connected to the vacuum drainage system in position over the area of leakage. UCTN – Unusual cases and technical notes E94