Role of inflammatory and nutritional markers in predicting complications after pancreaticoduodenectomy

Pancreaticoduodenectomy is associated with considerable morbidity and mortality rates. Early recognition of patients likely to develop severe postoperative complications will allow the timely commencement of a tailored approach. This study aimed to predict postoperative complications using inflammatory and nutritional markers measured early in the postoperative period. Patients who underwent pancreaticoduodenectomy between June 2019 and November 2020 were included in the study. Postoperative pancreatic fistula, delayed gastric emptying, and postoperative pancreatic hemorrhage were graded according to the International Study Group of Pancreatic Fistula and the International Study Group of Pancreatic Surgery. We also documented other complications such as wound infection, intra-abdominal collection, and nonsurgical complications. Nutritional and inflammatory markers were analyzed on postoperative days 1 and 3. Patients were followed up for 30 days or until discharge, depending on which was longer. Of the 58 enrolled patients, 51 were included in the study. The incidence of postoperative pancreatic fistula was 51% (clinically relevant postoperative pancreatic fistula 27.4%), delayed gastric emptying was 80.4% (clinically relevant delayed gastric emptying 43%), postoperative pancreatic hemorrhage was 3.9%, intra-abdominal collection was 23.5%, and wound infection was 29.4%. The median drain fluid interlukin-6 levels on postoperative day 1 and postoperative day 3 were significantly higher in patients developing clinically relevant postoperative pancreatic fistula than in those who did not develop clinically relevant postoperative pancreatic fistula on postoperative day 1 (211 [125, 425] fg/dL vs 99 [15, 170] fg/dL, [P = .045]) and on postoperative day 3 (110 [22, 28] fg/dL vs 10 [1.8, 45] fg/dL [P = .002]). Patients who tested negative for urine trypsinogen-2 on postoperative day 3 had a significantly lower probability of developing clinically relevant postoperative pancreatic fistula than those who tested positive (1 vs 24 [P < .001]). A model comprising both drain fluid interlukin-6 and urine trypsinogen-2 on postoperative day 3 definitively ruled out the occurrence of clinically relevant postoperative pancreatic fistula. Drain fluid interlukin-6 and urine trypsinogen-2 on postoperative day 3 ruled out the occurrence of clinically relevant postoperative pancreatic fistula.