Factors Predictive of Resolution in Pancreatic Duct Disruptions Treated Endoscopically: A 10 Year Single-Center Experience

Abstract

Factors Predictive of Resolution in Pancreatic Duct Disruptions Treated Endoscopically: A 10 Year Single-Center Experience Michel Kahaleh, Patrick Mcdonough, Thomas Rockoff, Jeffrey Tokar, Vanessa M. Shami, Reid B. Adams, Paul Yeaton Objective: The management of pancreatic duct disruption is well described, but no prospective study until now has confirmed factors associated with its resolution. Methods: Between Jan 95 and Sept 05, all patients with pancreatic ductal disruption were followed prospectively to complete resolution after endoscopic treatment. All patients underwent pancreatic sphincterotomy and every effort was made to place a stent if the disruption was crossed. All procedures were performed by 2 dedicated pancreatico-biliary endoscopists (MK and PY). Logistic regression analysis was performed on the following variables with regard to their ability to predict complete resolution (see table): age (!55 or R55 years), gender, etiology (alcoholic or not), chronicity (acute or chronic), pancreatic stenting, disruption type (complete versus partial), disruption location and enteral feeding(yes versus no). P value less than 0.05 were considered statistically significant. Results: 128 consecutive patients (85 male, 43 female), mean age: 51 G 15 years (range:18-88) were included. 103 patients presented with pancreatic fluid collections, 19 with ascites and 6 with pancreatic fistulas. Mean follow-up until resolution was 4.6 G 2.1 months (range:1-12) with enteral tube feeding placement in 78 patients (60%). 69 of the 128 patients (53%) had a stent placed with 63 of them (91%) bridging the disruption (9 from the minor ampulla). 116 patients (90%) had resolution of their pancreatic disruption, 7 patients underwent surgery, 3 required long term percutaneous drainage and 2 died from unrelated causes. Ten patients (8%) developed ERCP related complications: 3 self-limiting bleeding, two stent migrations retrieved endoscopically, two mild post ERCP pancreatitis and three cases of pneumoperitoneum managed conservatively. The factors statistically associated with complete resolution of the ductal disruption were stent placement and age R 55 years. Conclusion: Pancreatic ductal disruption can be effectively treated endoscopically. Resolution appears associated with successful stent placement bridging the disruption and age R 55 years old. Further studies are required to confirm this data.