Distinguishing between parenchymal and anastomotic leakage at duct-to-mucosa pancreatic reconstruction in pancreaticoduodenectomy

Aim: To distinguish anastomotic from parenchymal leakage at duct-to-mucosa reconstruction of the pancreatic remnant. Methods: We reviewed the charts of 68 pancreaticoduodenectomies performed between 5/2000 and 12/2005 with end-to-side duct-to-mucosa pancreatojejunostomy (PJ). The results of pancreatography, as well as peripancreatic drain volumes, and amylase levels were analyzed. Results: Of 68 pancreatojejunostomies, 48 had no leak by pancreatography and had low-drain amylase (normal); eight had no pancreatographic leak but had elevated drain amylase (parenchymal leak); and 12 had pancreatographic leak and elevated drain amylase (anastomotic leak). Although drain volumes in the parenchymal leak group were significantly elevated at postoperative day (POD) 4, no difference was found at POD 7. Drain amylase level was not significantly different at POD 4. In contrast, at POD 7, the anastomotic-leak group had significantly elevated drain amylase level compared with normal and parenchymal-leak groups (14158 ± 24083 IU/L vs 89 ± 139 IU/L and 1707 ± 1515 IU/L, respectively, P = 0.012). Conclusion: For pancreatic remnant reconstruction after pancreaticoduodenectomy, a combination of pancreatogram and peripancreatic drain amylase levels can be used to distinguish between parenchymal and anastomotic leakage at pancreatic remnant reconstruction.