Severe acute pancreatitis (SAP) is a disease of high (20-30%) mortality associated with prolonged hospitalization and excessive costs. Despite the major advances in our understanding of the pathophysiology of the disease, treatment remains supportive. The backbone of management is nutritional support as metabolic expenditure is excessively high and patients cannot eat for extended lengths of time . Because of concern that stimulation of the injured pancreas would exacerbate the disease process which is characterized by the premature activation of trypsinogen within the pancreas, the cornerstone of management has been pancreatic rest with intravenous feeding (TPN) which avoids pancreatic stimulation. However, the infective and metabolic complications of TPN have been shown to outweigh its benefits, and several prospective randomized comparative trials have demonstrated that patient outcome is better even with post-pyloric enteral feeding, which is stimulatory. Furthermore, 2 recent studies have collaborated studies in other populations of critically ill patients, showing that nasogastric (NG) feeding with a semielemental diet is as effective as post-pyloric feeding with the same dietary formula and does not increase the risk of aspiration despite the known impairmrnt of gastric emptying. This led to the recommendation that NG feeding should be used preferentially as it does not require expertize to start and to do. The concern remains that the mortality rate was not improved, raising the question whether NG feeding was better than no feeding. However, no feeding is not an option in SAP as unopposed protein catabolism would result in lifethreatening protein deficiency within 2 weeks. Exploratory studies of ours have suggested that feeding could be optimized by the placement of specialized double-lumen feeding tubes in the mid-jejunum (40cm past the ligament of Treitz) by transnasal endsocopic techniques to bypass the compressed upper Gl tract, to avoid pancreatic stimulation, and to stimulate the ileal brake - which experimentally has been shown to suppress acute pancreatitis. In the proposed study, we therefore plan to test the hypothesis that in comparison to simple NG feeding, skilled placement of DJ feeding tube systems hastens the resolution of disease because it is more effective in providing nutrition and does not exacerbate the disease process, thus leading to reduced morbidity, mortality and hospital costs. To recruit sufficient patients (n=114) to satisfy our statistical power calculations in a reasonable time period (5 years), we have formed a consortium with 8 leading national centers to conduct a multicenter clinical trial to achieve our goal.
