Utility of Drain Fluid Amylase Measurement on the First Postoperative Day after Pancreaticoduodenectomy Robert P. Sutcliffe • Narendra Battula • Ali Haque • Amir Ali • Parthi Srinivasan • Simon W. Atkinson • Mohamed Rela • Nigel D. Heaton • Andreas A. Prachalias Published online: 22 February 2012 Ó Socie ´te ´ Internationale de Chirurgie 2012 Abstract Background Early detection of pancreatic fistula (PF) may improve the outcome after pancreaticoduodenectomy, and exclusion of PF may allow earlier drain removal and accelerate recovery. The aim of the present study was to evaluate the relationship between drain fluid amylase on the first postoperative day (DFA 1 ) and PF. Patients and methods This work was designed as a pro- spective study and included patients undergoing pancrea- ticoduodenectomy in a single center. For each patient, DFA was measured on the first and fifth postoperative days, and PF was defined by drainage of amylase-rich fluid on the fifth postoperative day (DFA 5 [ 300 U/l). A cut-off value of DFA 1 was derived, which yielded sensitivity and neg- ative predictive value of 100% for predicting a PF. Results A total of 70 patients (47% male) who underwent pancreaticoduodenectomy (Whipple procedure: 37; pylo- rus-preserving procedure: 33) between April 2009 and March 2010 were included. Nine of those patients devel- oped a PF (grade A-2; B-5; C-2). There were two post- operative deaths (3%). The DFA 1 value significantly correlated with DFA 5 (Spearman rank coefficient 0.68; p \ 0.0001). The median DFA 1 of patients with a PF (6,205; range 357–23,391) was significantly higher than in patients without a PF (69; range 5–5,180; p = 0.01; unpaired t test). No patient with a PF had a DFA 1 B350 U/ l, compared to 48/61 patients (79%) without a PF. Using 350 U/l as a cut-off, a low DFA 1 excluded a PF with a sensitivity, specificity, positive and negative predictive values of 100, 79, 41, and 100%, respectively. Conclusions Drain fluid amylase on the DFA 1 after pancreaticoduodenectomy stratifies patients according to likelihood of developing a PF. Introduction Over the last three decades, perioperative mortality after pancreaticoduodenectomy has reduced from around 20% [1, 2] to less than 5% [3, 4]. Advances in surgical technique and perioperative management, and emergence of high volume specialist centers have all contributed to the observed improvements in outcome [5]. Postoperative pancreatic fistulas (PF) develop in 16–28% of patients [4, 6, 7], and the majority are minor and do not require intervention. Grade C PFs [8] are less common (2–5%) but are associated with significant morbidity, mortality, and prolonged hospitalization [6, 7]. PF has been defined as an output of amylase-rich fluid from an abdominal drain on or after the fifth postoperative day following pancreatico- duodenectomy [8]. In most centers, the surgical drain is left in situ until PF is excluded by a normal drain fluid amylase (DFA) value on day 5. Based on data from a number of randomized trials, routine drainage after several major abdominal procedures (e.g., colectomy and liver resection) is no longer considered necessary [9]. Indeed, prolonged drainage may increase the risk of infectious complications in some patients [9]. In a prospective non-randomized study of 104 patients who underwent pancreaticoduoden- ectomy, early removal of drains (day 4 vs. day 8) signifi- cantly reduced the risk of PF (3.6 vs. 23%; p = 0.0038) and intra-abdominal infections (7.7 vs. 38%; p = 0.0003) R. P. Sutcliffe (&) Á N. Battula Liver Unit, Queen Elizabeth Hospital, 3rd Floor, Nuffield House, Edgbaston, Birmingham B15 2TH, UK e-mail: robert.sutcliffe@uhb.nhs.uk A. Haque Á A. Ali Á P. Srinivasan Á S. W. Atkinson Á M. Rela Á N. D. Heaton Á A. A. Prachalias Institute of Liver Studies, King’s College Hospital, London, UK 123 World J Surg (2012) 36:879–883 DOI 10.1007/s00268-012-1460-0