OBJECTIVES: Amylase level in drains (ALD) has been proposed as a predictor for the development of a clinically significant pancreatic fistula (CS-PF) in patients undergoing a major pancreatic surgery. This study aimed to determine if the ALD in patients who developed a CS-PF after pancreatoduodenectomy is higher than that for patients with transient fistulae and to establish a threshold value as a predictor of a CS-PF. METHODS: From January 2002 to December 2012, all patients undergoing pancreatoduodenectomy were enrolled. At least 1 ALD measurement on postoperative day 3 was obtained. Pancreatic fistula (PF) was defined according to the International Study Group on Pancreatic Fistula. Both grade B and C PFs were considered as a CS-PF. We determined the cutoff value with a receiver operating characteristic curve. RESULTS: A total of 135 patients were enrolled. Pancreatic fistula was diagnosed in 36 cases (26.7%). The ALD median values were the following: PF grade A, 1809 U/L; PF grade B, 19,710 U/L; and PF grade C, 27,590 U/L. A drain amylase value of 2820 U/L was determined to be the cutoff for the development of a CS-PF. CONCLUSIONS: Patients with CS-PF have higher values of ALD than patients who developed a mild/transient fistula. An ALD higher than 2820 U/L identifies patients likely to present a CS-PF.
OBJECTIVES: Amylase level in drains (ALD) has been proposed as a predictor for the development of a clinically significant pancreatic fistula (CS-PF) in patients undergoing a major pancreatic surgery. This study aimed to determine if the ALD in patients who developed a CS-PF after pancreatoduodenectomy is higher than that for patients with transient fistulae and to establish a threshold value as a predictor of a CS-PF. METHODS: From January 2002 to December 2012, all patients undergoing pancreatoduodenectomy were enrolled. At least 1 ALD measurement on postoperative day 3 was obtained. Pancreatic fistula (PF) was defined according to the International Study Group on Pancreatic Fistula. Both grade B and C PFs were considered as a CS-PF. We determined the cutoff value with a receiver operating characteristic curve. RESULTS: A total of 135 patients were enrolled. Pancreatic fistula was diagnosed in 36 cases (26.7%). The ALD median values were the following: PF grade A, 1809 U/L; PF grade B, 19,710 U/L; and PF grade C, 27,590 U/L. A drain amylase value of 2820 U/L was determined to be the cutoff for the development of a CS-PF. CONCLUSIONS:Patients with CS-PF have higher values of ALD than patients who developed a mild/transient fistula. An ALD higher than 2820 U/L identifies patients likely to present a CS-PF.