Proposal of the novel predictive model for postoperative pancreatic fistula in distal pancreatectomy for pancreatic tumor based on preoperative imaging parameters and stapler handling.

Abstract

Distal pancreatectomy (DP) using linear staplers is widely performed; however, postoperative pancreatic fistulas (POPF) remain an issue. This study aimed to analyze preoperative risk factors for POPF and assess stapler handling. We retrospectively analyzed consecutive patients who underwent DP for pancreatic tumors using a linear stapler between 2014 and 2022. Preoperative measurements included pancreas-to-muscle signal intensity ratio (SIR) on fat-suppressed T1-weighted magnetic resonance imaging (MRI). The main outcome was clinically relevant POPF of the 2016 International Study Group of Pancreatic Fistulas definition. The predictive ability of the model was compared with the distal fistula risk score (D-FRS) by using the area under the receiver operating characteristic curve (AUROC). Among the 81 patients, POPF occurred in 31 (38.2%). Multivariate analysis identified computed tomography-measured pancreatic thickness (odds ratio [OR] 1.30, 95% confidence interval [CI] 1.05-1.61, p=0.009) and SIR on T1-weighted MRI (OR 6.85, 95% CI 1.71-27.4, p=0.002) as preoperative predictors. A novel preoperative model, "Thickness×MRI (TM)"-index, was established by multiplying these two variables. The TM-index exhibited the highest predictability preoperatively (AUROC 0.757, 95% CI 0.649-0.867). In the intraoperative variable analyses, TM-index (p<0.001), thin cartridge application (p=0.032), and short pre-firing compression (p=0.047) were identified as significant risk factors for POPF. The model's AUROC combined with these two stapler handling methods was higher than D-FRS (0.851 vs. 0.660, p=0.006). The novel preoperative model exhibited excellent predictability. Thick cartridge use and long pre-firing compression were protective factors against POPF. This model may facilitate preventive surgical strategy development to reduce POPF.