Prediction of post-operative pancreatic fistula in pancreaticoduodenectomy patients using pre-operative MRI: a pilot study

Background: Post-operative pancreatic fistula (POPF) is one of the most fearful complications which may occur after pancreaticoduodenectomy (PD). The methods used to predict POPF pre-operatively have not been studied in great detail. We analyzed correlation between various parameters related to PD including pre-operative magnetic resonance imaging (MRI) signal intensity (SI), pathology of pancreatic fibrosis and occurrence rates of POPF, and verified that MRI SI results could be the determining values for pre-operative prediction of POPF. Methods: From January 2005 to August 2006, we retrospectively examined 43 cases of PDs by reviewing abdominal MRI findings, degree of fibrosis of remnant pancreatic stump, and other surgery-related parameters. Results: POPF encountered in PD were 11 cases (25.6%). Operation time and degree of fibrosis of remnant pancreatic cut surface were related to POPF (P = 0.030, P = 0.010). The pancreas-liver SI ratio (PLSI) between fistula group and no fistula group was -0.0009 +/- 0.2 and -0.1297 +/- 0.2, respectively (P = 0.0004). The pancreas-spleen SI ratio (PSSI) in each group was 0.423 +/- 0.25 and 0.288 +/- 0.32, respectively (P = 0.014). Using quantitative analysis, the SI ratios were 1.27 and 0.66 in each group (P = 0.013). Conclusions: When analyzing the results of POPF in 43 patients who underwent PD, PLSI, PSSI and qualitative analysis, fistula group differed significantly from no fistula group. Using these results, it will be helpful for us to predict the occurrence of POPF pre-operatively using MRI in PD patients.