Prediction of Pancreatic Fistula After Distal Pancreatectomy Based on Cross-Sectional Images

Background Postoperative pancreatic fistula (POPF) is one of the most common and clinically relevant complications after distal pancreatectomy (DP), occurring in 5-40% of patients. Determining risk factors for this complication may aid in its prevention. This study sought to predict the development of POPF after DP preoperatively and objectively based on radiologic findings. Methods This study included 60 patients who underwent DP using a stapler for pancreatic division between June 2011 and January 2013. Fatty infiltration, apparent diffusion coefficients (ADC) on preoperative MRI, pathologic fat, and fibrosis were measured. Pancreatic thickness and cross-sectional area of the pancreas stump on CT scan were also measured. Results Mean patient age was 60.5 years, 26 patients (46.3%) had pancreatic cancer and 20 (33.3%) underwent laparoscopic surgery. Clinically relevant POPF was observed in 12 patients (20.0%). Linear regression analysis showed a significant correlation between fat quantification on MRI and pathologic fat (pathologic fat = 1.978 x MR fat -6.393, p < 0.001, R-2 = 0.777). Univariate analysis showed that <= 8% fat on MRI (p = 0.040), <= 5% pathologic fat (p = 0.002), ADC <= 1.3 9 x 10(-3) mm(2)/s (p = 0.020), thicker pancreas (p = 0.007), and wider cross-sectional area of the pancreas (p = 0.013) were significantly associated with clinically relevant POPF after DP. Multivariate analysis revealed that pancreas thickness > 17.6 mm [odds ratio (OR) 6.532, p = 0.064] and cross-sectional area > 377 mm(2) (OR 12.676, p = 0.052) were marginally related to clinically relevant POPF. Conclusions Pancreatic thickness and cross-sectional area of the transected surface of the pancreas are marginally significant risk factors for POPF development after DP. Measuring pancreatic thickness and cross-sectional area can be a promising tool for the preoperative prediction of POPF.