PURPOSE: Hepatobiliary scintigraphy is commonly performed on liver transplant patients to exclude biliary complications such as obstruction or leak. Biliary reconstruction in these patients is performed using either a direct duct-to-duct anastomosis (choledochocholedochostomy) or a biliary-enteric anastomosis (Roux-en-Y hepaticojejunostomy). The authors have observed a finding in scans performed on the latter group of patients that often simulates a bile leak. They postulate this finding to represent retention of activity in the blind end of the Roux limb (the "blind end sign"). The purpose of this study was to determine characteristics of this blind end activity, which may help differentiate it from true bile leaks. METHODS: Two nuclear medicine physicians reviewed 36 Tc-99m mebrofenin hepatobiliary scans (31 patients). Biliary reconstruction was performed in 25 patients with a Roux-en-Y hepaticojejunostomy, and in 6 patients with a choledochocholedochostomy. Each scan was evaluated regarding whether there was focal, persistent retention of activity in the anastomotic region, and, if so, various features of the activity were noted. Activity was concluded to represent a true bile leak based on subsequent surgical confirmation, and if there was no such confirmation, the activity was concluded to represent blind end retention in patients with Roux-en-Y hepaticojejunostomies. RESULTS: There were 8 total examples of blind end activity and 3 examples of bile leak activity. Useful discriminators between the 2 types of activity included the following: fluctuation in size and/or intensity with time (8 of 8 blind end cases vs. 0 of 3 bile leak cases); size comparable with bowel (8 of 8 vs. 1 of 3); tubular (4 of 8 vs. 0 of 3), round (4 of 8 vs. 1 of 3), or irregular (0 of 8 vs. 2 of 3) shape; and movement into bowel (3 of 8 vs. 0 of 3). Delayed imaging was performed in 3 cases at 1.5 hours, but was only conclusive in 1 case in which diffuse extravasation was seen consistent with bile leak. CONCLUSION: Retention of activity in the blind end of the Roux limb in patients with Roux-en-Y hepaticojejunostomies may simulate bile leak, but certain typical features, as well as delayed imaging, may aid in distinguishing the two.
PURPOSE:Hepatobiliary scintigraphy is commonly performed on liver transplant patients to exclude biliary complications such as obstruction or leak. Biliary reconstruction in these patients is performed using either a direct duct-to-duct anastomosis (choledochocholedochostomy) or a biliary-enteric anastomosis (Roux-en-Y hepaticojejunostomy). The authors have observed a finding in scans performed on the latter group of patients that often simulates a bile leak. They postulate this finding to represent retention of activity in the blind end of the Roux limb (the "blind end sign"). The purpose of this study was to determine characteristics of this blind end activity, which may help differentiate it from true bile leaks. METHODS: Two nuclear medicine physicians reviewed 36 Tc-99m mebrofeninhepatobiliary scans (31 patients). Biliary reconstruction was performed in 25 patients with a Roux-en-Y hepaticojejunostomy, and in 6 patients with a choledochocholedochostomy. Each scan was evaluated regarding whether there was focal, persistent retention of activity in the anastomotic region, and, if so, various features of the activity were noted. Activity was concluded to represent a true bile leak based on subsequent surgical confirmation, and if there was no such confirmation, the activity was concluded to represent blind end retention in patients with Roux-en-Y hepaticojejunostomies. RESULTS: There were 8 total examples of blind end activity and 3 examples of bile leak activity. Useful discriminators between the 2 types of activity included the following: fluctuation in size and/or intensity with time (8 of 8 blind end cases vs. 0 of 3 bile leak cases); size comparable with bowel (8 of 8 vs. 1 of 3); tubular (4 of 8 vs. 0 of 3), round (4 of 8 vs. 1 of 3), or irregular (0 of 8 vs. 2 of 3) shape; and movement into bowel (3 of 8 vs. 0 of 3). Delayed imaging was performed in 3 cases at 1.5 hours, but was only conclusive in 1 case in which diffuse extravasation was seen consistent with bile leak. CONCLUSION: Retention of activity in the blind end of the Roux limb in patients with Roux-en-Y hepaticojejunostomies may simulate bile leak, but certain typical features, as well as delayed imaging, may aid in distinguishing the two.