Effects of antecolic versus retrocolic reconstruction for gastro/duodenojejunostomy on delayed gastric emptying after pancreatoduodenectomy: a systematic review and meta-analysis

Background: Delayed gastric emptying (DGE) is a relatively common complication after pancreatoduodenectomy (PD). The aim of this study was to determine whether DGE is affected by antecolic or retrocolic reconstruction for gastro/duodenojejunostomy after PD. Methods: A literature search was performed of the MEDLINE (PubMed), Ovid SP, ISI Web of Knowledge, EMBASE, and Cochrane databases to identify randomized controlled trials (RCTs) and clinical observational studies related to this topic from January 1995 to November 2014. Pooled odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for categorical outcomes, and mean differences (MD) using fixed-effect and random-effects models were calculated for the meta-analysis. Results: Fourteen studies including 1969 patients met the inclusion criteria. Six studies were RCTs, and eight studies were clinical observational studies. DGE was less common in the antecolic reconstruction group than in the retrocolic reconstruction group (OR = 0.24 [0.12 -0.48], P < 0.0001). Postoperative days to start solid foods (MD = -3.67 d [-5.10 to -2.33], P < 0.00001) and length of hospital stay (MD = -2.90 d [-5.36 to -2.33], P < 0.00001) were also significantly in favor of the antecolic reconstruction group. There was no difference in the incidence of pancreatic fistula, intra-abdominal fluid collection or abscess, biliary fistula, or mortality. However, in the subgroup analyses, using the data of six RCTs or seven studies according to the International Study Group of Pancreatic Surgery definition, there was no significant difference in the incidence of DGE. Conclusions: Antecolic reconstruction for gastro/duodenojejunostomy does not seem to offer an advantage over retrocolic reconstruction with respect to DGE after PD. (C) 2016 Elsevier Inc. All rights reserved.