A simple preoperative stratification tool predicting the risk of postoperative pancreatic fistula after pancreatoduodenectomy

Background: Postoperative pancreatic fistula (POPF) after pancreatoduodenectomy (PD) can be associated with severe postoperative morbidity. This study aims to develop a preoperative POPF risk calculator that can be easily implemented in clinical routine.

Methods: Patients undergoing PD were identified from a prospectively-maintained database. A total of 11 preoperative baseline and CT-based radiological parameters were used in a binominal logistic regression model. Parameters remaining predictive for grade B/C POPF were entered into the risk calculator and diagnostic accuracy measures and ROC curves were calculated for a training and a test patient cohort. The risk calculator was transformed into a simple nomogram.

Results: A total of 242 patients undergoing PD in the period from 2012 to 2018 were included. CT-imaging-based maximum main pancreatic duct (MPD) diameter (p = 0.047), CT-imaging-based pancreatic gland diameter at the anticipated resection margin (p = 0.002) and gender (p = 0.058) were the parameters most predictive for grade B/C POPF. Based on these parameters, a risk calculator was developed to identify patients at high risk of developing grade B/C POPF. In a training cohort of PD patients this risk calculator was associated with an AUC of 0.808 (95%CI 0.726-0.874) and an AUC of 0.756 (95%CI 0.669-0-830) in the independent test cohort. A nomogram applicable as a visual risk scale for quick assessment of POPF grade B/C risk was developed.

Conclusion: The preoperative POPF risk calculator provides a simple tool to stratify patients planned for PD according to the risk of developing postoperative grade B/C POPF. The nomogram visual risk scale can be easily integrated into clinical routine and may be a valuable model to select patients for POPF-preventive therapy or as a stratification tool for clinical trials.