Pancreatic anastomoses after pancreaticoduodenectomy: do we need further studies?

Pancreatic anastomotic leak is the single most important factor responsible for the considerable morbidity and mortality associated with pancreaticoduodenectomy. Management of the pancreatic remnant is controversially discussed, reflecting the complexity of anastomosing a pancreas of different textures to the digestive tract. A number of studies evaluating diverse options have often provided conflicting conclusions. This information is confusing particularly to those surgeons outside of large-volume centers with broad experience and to general surgeons who perform pancreatic surgery. A PubMed search with the key words pancreaticoduodenectomy, pancreatic anastomosis, pancreaticojejunostomy, pancreaticogastrostomy, and pancreatic fistula was performed. Major series of pancreatic anastomosis published between 1990 and 2002 were studied from diverse centers worldwide. Their results with regard to pancreatic fistula, morbidity, and mortality were documented. Nine series of pancreaticojejunostomy and seven series of pancreaticogastrostomy were evaluated. Eight comparative studies evaluating the two techniques were also analyzed. A single randomized controlled trial was identified among these comparative studies. Equally good results were observed with the two techniques. Other uncommon methods of management of the pancreatic remnant (duct occlusion and ligation) were also evaluated. Pancreaticojejunostomy followed by pancreaticogastrostomy are the most favored techniques. A duct-to-mucosa anastomosis is preferred over other methods. Fistula rates of less than 5%-10% should be the standard irrespective of the technique used. Unlike in the past, mortality can be reduced even in the event of an anastomotic dehiscence, and this aspect is primarily dependent on a meticulous anastomosis based on sound surgical principles rather than the method per se. Based on the information accumulated, adherence to these specific principles could ensure a safe and reliable pancreatic anastomosis with mimimal morbidity and mortality after pancreaticoduodenectomy, even in the hands of general surgeons operating outside high-volume centers.