Comparison of lipase and amylase for diagnosing post‐operative pancreatic fistulae

BACKGROUND: To ascertain if post-operative drain fluid lipase is superior to amylase in routinely detecting clinically significant post-operative pancreatic fistulae (POPF). METHOD: Between January 2015 and March 2016 data were collected on all patients undergoing pancreatic surgery at a regional referral centre. Routine drain fluid analysis was performed on post-operative patients as part of a locally defined enhanced recovery protocol. POPF was diagnosed in accordance with the recently updated International Study Group of Pancreatic Surgery guidance. RESULTS: During the study period, there were 68 pancreatic resections. The median age was 69.1 years. A total of 11 (15.9%) patients developed clinically significant POPF (nine type B and two type C). The median drain amylase result in patients with type B or C leak was 532 IU/L (interquartile range (IQR) 264-833). This was significantly higher than those without a clinical fistula (median 38, IQR 15-376, P = 0.012). The median drain lipase result was 1504 IU/L (IQR 746-2236). This was significantly higher than those without a clinical fistula (median 57, IQR 13-1277, P = 0.012). Fluid amylase had a sensitivity of 81.8% and specificity of 69.2%; fluid lipase had a sensitivity of 91% and specificity of 64.9%. CONCLUSION: Our experience suggests drain fluid amylase or lipase results are not sufficiently sensitive or specific to reassure clinicians and rule out clinically significant POPF. However, if biochemical tests are used to aid decision-making, then lipase is a more sensitive biochemical marker than amylase for the routine detection of clinically significant POPF.