Measurement of Pancreatic Fat by Magnetic Resonance Imaging Predicting the Occurrence of Pancreatic Fistula After Pancreatoduodenectomy

Abstract

Objective: The purpose of this study was to determine whether patients who develop a pancreatic fistula after pancreatoduodenectomy are more likely to have higher pancreatic fat levels than matched controls and if so, to investigate whether preoperative dual gradient-echo magnetic resonance (MR) imaging can be used to measure pancreatic fat and predict the development of postoperative pancreatic fistula. Summary Background Data: Pancreatic fistula is a major complication and its most frequently reported risk factors tend to be anatomic features of the pancreatic remnant, such as a soft pancreatic texture. Methods: Between January 2007 and September 2007, a total of 96 cases of pancreatoduodenectomy were performed at Seoul National University Hospital. Of total, 20 patients (20.8%) who developed a pancreatic fistula were carefully matched for multiple parameters including age, gender, pancreatic pathology, surgeon, and type of operation with 20 control patients who did not develop a pancreatic fistula. In the pancreatic fistula group, 9 patients (45%) had a grade A fistula and 11 (55%) grade B fistula. Degrees of pancreatic fatty infiltration and fibrosis were assessed quantitatively. In-phase and opposed-phase images were obtained by dual-gradient-echo MR imaging. Percentage decreases in pancreatic signal intensity on opposed-phase images relative to those on in-phase images were calculated and defined as relative signal intensity decreases (RSID). Results: More patients in the pancreatic fistula group had a soft pancreatic texture or a smaller pancreatic duct and total fat and RSID were significantly higher. In soft pancreatic texture group, intralobular, interlobular, and total fat were significantly elevated. Furthermore, pancreatic fat levels were found to be correlated positively with RSID (P = 0.013). RSID was correlated with total pancreatic fat and when an RSID criterion more than 7.032 was used, pancreatic fistula could be predicted as 72.7% sensitivity and 75.9% specificity. Conclusion: Our findings suggest that increased pancreatic fat is a risk factor of postoperative pancreatic fistula. Preoperative measurements of pancreatic fat by MRI offer a noninvasive predicting the occurrence of pancreatic fistula.