Abstract
PURPOSES: To establish a scoring model for the risk of postoperative pancreatic fistula (POPF) following pancreatoduodenectomy (PD). METHODS: PD Patients from 7 institutions in 2 independent sets: developmental (n = 457) and validation cohort (n = 152) were retrospectively enrolled and analyzed. Pancreatic Fibrosis (PF) and Pancreatic Steatosis (PS) were assessed by pathological examination of the pancreatic stump. RESULTS: Stepwise univariate and multivariate analysis indicated that pancreatic duct diameter ≤ 3 mm, increased PS and decreased PF were independent risk factors for POPF and Clinically Relevant Postoperative Pancreatic Fistula (CR-POPF). Based on the relative weight and odds ratio of each factor in the POPF, a simplified scoring model was developed. And patients were stratified into high-risk group (22~28 points), medium-risk group (15~21 points) and low-risk group (8~14 points). The receiver operating characteristic curve demonstrated that the Area under the curve for the predictive model was 0.868 and 0.887 in the model design group and the external validation group. CONCLUSIONS: This study establishes a simplified scoring model based on accurately and quantitatively measuring the PS, PF and pancreatic duct diameter. The scoring model accurately predicted the risk of POPF.