ERCP Evaluation and Treatment for Pancreaticobiliary Complications Following Pancreaticoduodenectomy (PDE/Whipple's Resection)

Abstract

Background: Ductal anastomotic stricturing not infrequently follows PDE, especially with long-term survival, and may be approached endoscopically. The success and outcome of endoscopic intervention in this setting has been little reported. Patients (Pts) and Methods: From 1/1993 to 6/2006, 39 pts were identified in a prospective ERCP database as having undergone a total of 65 attempted ERCPs in PDE pts. Pts had undergone PDE a median of 39.5 months earlier (1-396 months) and indications for the surgery included chronic pancreatitis (n = 9), pancreatic cancer (n = 9), ampullary adenoma (n = 8), cholangioca (n = 5), neuroendocrine tumor (n = 4), or other (n = 4). The indications for ERCP were biliary (jaundice, colic, cholangitis, abnormal LFTs) in 23/39 (59%) and pancreatic (pancreatitis, pancreatic-type pain, pseudocyst, steatorrhea) in 16/39 (41%). ERCPs were performed with standard side-viewing endoscopes, pediatric colonoscopes, or prototype oblique-viewing enteroscopes and standard or special long-length accessories. Results: Success for planned interventions was achieved in 29/43 attempts (67%) for biliary indications and 10/21 attempts (48%) for pancreatic indications. Total biliary interventions included stent placement (n = 17), balloon dilation (n = 9), stone extraction (n = 7) and diathermic stricture release (n = 6). Total pancreatic interventions included stent placement (n = 8), balloon dilation (n = 3), diathermic stricture release (n = 1) and pseudocyst drainage (n = 1). Successful cannulation of the desired duct depended on endoscopic access to the afferent jejunal limb and was independent of the type of PDE (conventional with antrectomy (n = 21) or with pylorus preservation (n = 18)). In follow-up, of the 24 pts who underwent successful endoscopic intervention, 10 pts (42%) required an average of 2.9 (range 2-7) re-interventions. Minimum follow-up of 6 months after the final intervention was determined in 22/24 pts (92%) who had undergone successful cannulation and endoscopic treatment. 18 of these 22 (81%) pts were alive and 14/22 (64%) pts were asymptomatic. Four pts continued to have symptoms despite patent anastomoses at endoscopy. Four pts have died from progressive cancer. Of the 15 unsuccessful ERCP pts, 8 have undergone additional procedures including PTC (3 pts) or surgical revision (5 pts). Two pts have died and 5 continue to be followed. Conclusions: Endoscopic intervention in the setting of Whipple's resection for a variety of indications can often be achieved with positive outcome, especially for biliary indications. The approach to and the identification of the pancreatic anastomosis is technically challenging resulting in failure in >50%.