Back to 2017 Posters
GASTROINTESTINAL FISTULA IN NECROTIZING PANCREATITIS- A SINGLE TERTIARY CARE CENTER EXPERIENCE
Guru Trikudanathan*1, Michael Schnaus2, Pierre Tawfik2, Alyssa Liubakka2, Mustafa A. Arain1, Rajeev Attam1, J. Shawn Mallery1, Stuart K. Amateau1, Martin L. Freeman1 1Gastroenterology, University of Minnesota, Minneapolis, MN; 2Internal Medicine, University of Minnesota, Minneapolis, MN
Background and Aim: Gastrointestinal (GI) fistula is a well-recognized complication of necrotizing pancreatitis (NP). To report the clinical presentation, course, outcome and management of all necrotizing pancreatitis patients with GI fistulae in our institution. Methods: A retrospective analysis of all NP patients managed at our institution between 2010-16 was performed from a prospectively maintained database for the occurrence of GI fistulae. Demographic characteristics, clinical presentation, time interval between pancreatitis and detection of fistula, management and outcome was evaluated. Results: GI fistulization was detected in 61 (18%) of the 330 NP managed during the study period. Among these 44 (71%) were males with a median (range) age was 56 (23-84 years) with majority 51 (83%) whites. Etiology for AP was biliary in 31 (50%), alcohol induced in 14 (23%), hypertriglyceridemia induced in 2 (4%), post ERCP in 2 (3%) and idiopathic in 12 (20%). Fistula developed spontaneously in 22 (36%) pts while 39 (64%) had prior surgical or percutaneous intervention. Presentations included abdominal pain in 42 (69%), fever(sepsis) in 34 (55%), persistent cutaneous discharge in 24 (39%), gastrointestinal bleeding in 13 (21%) and shortness of breath in 3 (5%). Endoscopic visualization of fistula was possible in 34 (56%). Location of the fistula was stomach in 6 (8%), duodenum in 30 (49%), colon in 10 (17%), pleural in 3 (5%) and cutaneous in 13 (21%). Spontaneous fistulous tract from duodenum into necrotic collection was dilated and stent was placed for endoscopic drainage in 21 (34%). Other endoscopic closure modalities were successful in 2 (3%). Surgical management included fistula take down in 9 (15%), bowel resection in 5 (8%) and thoracotomy in 1 (2%) patient with a pleural fistula. Fistula closure was successful in 35 (57%). Overall all-cause mortality NP patients with GI fistulae was 8 (10%). Conclusion GI fistulae occurred in 18% of our NP patients, with majority having a history of prior surgical or percutaneous intervention. Fistulae closure was successful in a majority of pts with an overall mortality of 10%.
Back to 2017 Posters
|