Hypoattenuated Area at Pancreatojejunostomy Site After Pancreatoduodenectomy Predicts Grade B/C Postoperative Pancreatic Fistula

Background After pancreatojejunostomy for pancreatic head resection, contrast-enhanced computed tomography sometimes reveals a hypoattenuated area at the pancreatojejunostomy site. The clinical impact of this hypoattenuated area, in terms of postoperative pancreatic fistula, is unknown. Methods Among 309 patients undergoing pancreatic head resection with pancreatojejunostomy reconstruction at Osaka University Hospital from 2012 to 2020, 105 exhibited drain amylase levels over 3× the upper limit of normal. According to contrast-enhanced computed tomography on postoperative days 3–14, these patients were divided into two groups—evident hypoattenuated area group (≥ 5 mm; n = 46) and subtle hypoattenuated area group (< 5 mm; n = 59)—and evaluated for clinically relevant pancreatic fistula. Results Clinically relevant pancreatic fistula was significantly more common with the evident hypoattenuated group (80.4%; 37/46) than the subtle hypoattenuated group (40.7%; 24/59) ( P < 0.0001). Multivariate analysis identified four factors related to clinically relevant pancreatic fistula formation: male sex ( P = 0.0230), main pancreatic duct diameter < 3 mm ( P = 0.0007), operative time > 475 min ( P = 0.0161), and evident hypoattenuated area group ( P = 0.0050). Hypoattenuated area ≥ 5 mm was associated with clinically relevant pancreatic fistula (60.7% sensitivity; 79.6% specificity). Conclusion Evident hypoattenuated area on postoperative contrast-enhanced computed tomography was an independent risk factor for clinically relevant pancreatic fistula. Patients experiencing increased drain amylase levels may benefit from contrast-enhanced computed tomography assessment to predict clinically relevant pancreatic fistula formation.