Back to Program | 2010 Program and Abstracts Overview | 2010 Posters

Background: Delayed gastric emptying (DGE) is a frequent and bothersome complication after pancreatoduodenectomy. Some authors suggest that an antecolic route of the gastroenteric anastomosis (duodenojejunostomy, DJ or gastrojejunostomy, GJ) lowers the incidence of DGE, compared to a retrocolic route. In our institution, a retrocolic route has been routinely used until 2005, after which an antecolic route became more frequent.Aim: To investigate the relation between the route of gastroenteric anastomosis and the incidence of DGE after pancreatoduodenectomy.Methods: In a consecutive series of 203 patients from our prospective pancreatoduodenectomy database, the route of gastroenteric anastomosis was established by reviewing operation reports. Hospital course and follow-up were prospectively recorded. Patients with antecolic DJ or GJ were compared to patients with retrocolic DJ or GJ. Main outcome measure was the incidence of DGE according to the criteria of the International Study Group of Pancreatic Surgery. Secondary outcome measures were other complications and length of hospital stay. Results: In 47 patients the route of gastroenteric anastomosis could not be determined. Two patients were excluded because they had Roux-en-Y reconstruction. In the remaining 154 patients, 77 had a retrocolic anastomosis and 77 had an antecolic anastomosis. In the retrocolic group, DGE occurred in 58% of patients (25% grade A, 17% grade B and 17% grade C). In the antecolic group, 52% had DGE (21% grade A, 16% grade B and 16% grade C). This difference was not significant. ‘Primary’ DGE of any grade (not due to other intra-abdominal complications) occurred in 36% of the retrocolic group and 20% of the antecolic group (p 0.02). ‘Primary’ clinically relevant DGE (grade B or C) occurred in 18% and 10%, respectively (p 0.17). There was no difference in need for (par)enteral nutritional support, other complications, hospital mortality or length of hospital stay.Conclusions: The route of DJ or GJ had no influence on the overall postoperative incidence of DGE. Clinically relevant DGE (overall and ‘primary’) was not different between the retrocolic and antecolic group. ‘Primary’ DGE (any grade) was more frequent in the retrocolic group, mainly due to a higher incidence of DGE grade A. The preferred route for gastroenteric anastomosis in pancreatoduodenectomy remains to be confirmed in a well-powered randomized controlled trial.