Meta-analysis Meta-analysis of pancreaticojejunostomy versus pancreaticogastrostomy reconstruction after pancreaticoduodenectomy A. McKay 1 , S. Mackenzie 1 , F. R. Sutherland 1 , O. F. Bathe 1 , C. Doig 2 , J. Dort 1,2 , C. M. Vollmer Jr 3 and E. Dixon 1 Departments of 1 Surgery and 2 Community Health Sciences, University of Calgary, Calgary, Canada and 3 Department of Surgery, Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA Correspondence to: Dr E. Dixon, Department of Surgery, University of Calgary, Tom Baker Cancer Centre, 1331 29th Street NW, Calgary, Alberta, Canada T2N 4N2 (e-mail: Elijah.Dixon@CalgaryHealthRegion.ca) Background: Pancreaticoduodenectomy is the primary treatment for periampullary cancer. Associated morbidity is high and often related to pancreatic anastomotic failure. This paper compares rates of pancreatic fistula, morbidity and mortality after pancreaticoduodenectomy in patients having reconstruction by pancreaticogastrostomy with those in patients having reconstruction by pancreaticojejunostomy. Methods: A meta-analysis was performed of all large cohort and randomized controlled trials carried out since 1990. Results: Eleven articles were identified for inclusion: one prospective randomized trial, two non- randomized prospective trials and eight observational cohort studies. The meta-analysis revealed a higher rate of pancreatic fistula associated with pancreaticojejunostomy reconstruction (relative risk (RR) 2·62 (95 per cent confidence interval (c.i.) 1·91 to 3·60)). A higher overall morbidity rate was also demonstrated in this group (RR 1·43 (95 per cent c.i. 1·26 to 1·61)), as was a higher mortality rate (RR 2·51 (95 per cent c.i. 1·61 to 3·91)). Conclusion: Current literature suggests that the safer means of pancreatic reconstruction after pancreaticoduodenectomy is pancreaticogastrostomy, but much of the evidence comes from observational cohort study data. Paper accepted 20 February 2006 Published online in Wiley InterScience (www.bjs.co.uk). DOI: 10.1002/bjs.5407 Introduction Pancreaticoduodenectomy is the primary means of treat- ing periampullary cancer. Associated mortality rates have dropped dramatically in the past decade, but morbid- ity rates remain high 1,2 . Despite a significant amount of research, the pancreas to jejunum connection con- tinues to be the major cause of this morbidity 3–13 , and efforts to solve this problem have included performing total pancreatectomy, oversewing the end of the pancreas and injecting the pancreatic duct with silicone 14–16 . Many techniques are used to connect the pancreatic stump with The Editors have satisfied themselves that all authors have contributed significantly to this publication the jejunum but, despite some success, problems remain, especially with a soft pancreas gland and small duct 17 . Pancreatic juice entering the jejunum becomes activated with enterokinase, and any leakage of the connection includes succus entericus and bile 6,14,17 . Obstruction of the jejunal limb can result in positive pressure on the pancreaticojejunal anastomosis, increasing the chance of leakage. Oedema of the jejunal limb may also increase the likelihood of leakage and subsequent morbidity. The stomach offers an obvious alternative for making the pancreas – enteric connection. It is thick walled, well vascu- larized and takes sutures well. Furthermore, the acidic envi- ronment may prevent activation of proteolytic enzymes, and negative pressure can be maintained with a nasogastric tube 6 . Copyright  2006 British Journal of Surgery Society Ltd British Journal of Surgery 2006; 93: 929–936 Published by John Wiley & Sons Ltd