Meta-analysis of complication rates for single-loop versus dual-loop (Roux-en-Y) with isolated pancreaticojejunostomy reconstruction after pancreaticoduodenectomy.

BACKGROUND: Postoperative pancreatic fistula is one of the most important and potentially severe complications after partial pancreaticoduodenectomy. In this context, the reduction of postoperative pancreatic fistula by means of a dual-loop (Roux-en-Y) reconstruction with isolation of the pancreaticojejunostomy from biliary drainage has been evaluated in several studies. This systematic review and meta-analysis summarizes evidence of effectiveness and safety of the isolation of the pancreaticojejunostomy compared with conventional single-loop reconstruction.

METHODS: Randomized clinical trials (RCTs) and controlled clinical trials (CCTs) comparing outcomes of dual-loop reconstruction with isolated pancreaticojejunostomy and single-loop reconstruction were searched according to PRISMA guidelines. Random-effects meta-analyses were performed and the results presented as weighted risk ratios or mean differences with their corresponding 95 per cent c.i.

RESULTS: Of 83 trials screened for eligibility, three RCTs and four CCTs including a total of 802 patients were finally included. Quantitative synthesis showed no significant statistical difference between the two procedures regarding postoperative pancreatic fistula, delayed gastric emptying, haemorrhage, intra-abdominal fluid collection or abscess, bile leakage, wound infection, pneumonia, overall morbidity, mortality, reinterventions, reoperations, perioperative blood loss and length of hospital stay. Duration of surgery was significantly longer in patients undergoing dual-loop reconstruction.

CONCLUSION: Dual-loop (Roux-en-Y) reconstruction with isolated pancreaticojejunostomy after partial pancreaticoduodenectomy is not superior to single-loop reconstruction regarding pancreatic fistula rate or other relevant outcomes. Additional superiority trials are therefore not warranted, although a high-quality trial may be justified to prove equivalence or non-inferiority.