Systematic review and meta-analysis of outcomes after intraoperative pancreatic duct stent placement during pancreaticoduodenectomy

Abstract

Postoperative pancreatic leakage after pancreaticoduodenectomy is often serious. Although some studies have suggested that stenting the anastomosis can reduce the incidence of this complication, the value of stenting in the setting of pancreaticoduodenectomy remains unclear. Studies comparing outcomes of stent versus no stent, and internal versus external stent placement for pancreaticoduodenectomy were eligible for inclusion. Pooled odds ratios (ORs) with 95 per cent confidence intervals were calculated using fixed- or random-effects models. From a search of the literature published between January 1973 and September 2011, five randomized clinical trials (RCTs) and 11 non-randomized observational clinical studies (OCS) involving 1726 patients were selected for inclusion in this review. Meta-analysis of RCTs revealed that placing a stent in the pancreatic duct did not reduce the incidence of postoperative pancreatic fistula. External stents had no advantage over internal stents in terms of clinical outcome. Subgroup analyses revealed that use of an external stent significantly reduced the incidence of pancreatic fistula (RCTs: OR 0·42, 0·24 to 0·76, P = 0·004; OCS: OR 0·43, 0·27 to 0·68, P < 0·001), delayed gastric emptying (RCTs: OR 0·41, 0·19 to 0·87, P = 0·02) and postoperative morbidity (RCTs: OR 0·55, 0·34 to 0·89, P = 0·02) compared with no stent. Pancreatic duct stenting did not reduce the incidence of pancreatic fistula and other complications in pancreaticoduodenectomy compared with no stenting. Although no difference was found between external and internal stents in terms of efficacy, external stents seemed to reduce the incidence of pancreatic fistula compared with control.