Endoscopic management of pancreatic pseudocyst by EUS localization of puncture site and balloon dilatation of fistulae

Endoscopic drainage of pancreatic pseudocysts (PPC) is a minimally invasive and inexpensive alternative to surgical management. Risks of endoscopic therapy relate to improper fistula placement (hemorrhage, failed access) or fistula size (premature closure, inadequate drainage, infection). Aim: To assess endoscopic PPC drainage at our institution utilizing EUS for cyst and puncture site localization and wire-guided balloon dilation of the fistula track. Methods: We reviewed the procedure and subsequent course of patients referred for endoscopic pseudocyst drainage. EUS was used to assess cyst wall and contents and to select an optimal site for puncture as judged by close apposition of PPC and gut wall and absence of vascular structures. Non-EUS endoscopic management included fistulotomy under fluoroscopic guidance using a 0.35 inch monopolar fistulatome, wire and non-wire guided ballon dilatation, cyst lavage, nasocystic drainage and/or stent placement and prophylactic antibiotics. Results: Of 13 patients assessed for endoscopic drainage, 2 were rejected solely on the basis of EUS findings: prominent gastric varices in one and lack of adequate cyst-stomach apposition in the other. Eleven patients were treated, 10 by cyst-gastrostomy and 1 by cyst-duodenostomy. Mean maximal cyst diameter was 9 cm (range: 5-12 cm). Pseudocysts were due to acute pancreatitis in 5 patients, chronic pancreatitis in 7 and malignancy in 1. Puncture and initial stent placement were successful in all patients. Fistulae were dilated in 8/11 patients using balloons to maximal diameter of 8mm (2), 12mm (4) and 18mm (2). Lavage of the cavity was routinely performed. No significant hemorrhage occurred. Two patients had initial naso-cystic drainage followed by dual stents after several days; 9 had plastic stents from the outset. Patency was maintained using a single 11.5F plastic stent in 1 patient, dual stents in 7 patients and 3 stents in 2 patients. Stents were removed after a median period of 8 weeks. One patient who had not received ongoing antibiotic prophylaxis developed sepsis requiring partial pancreatectomy after 15 days. At follow-up, PPC had resolved in 9 patients and decreased in size (and become asymptomatic) in 1. Conclusions: EUS is useful for assessment of anatomic suitability for drainage and for selection of optimal puncture sites. Balloon dilatation is an effective and safe method of enlarging the fistula track to ensure adequate drainage. Endoscopic drainage of symptomatic pseudocysts by this method was successful in 10/11 patients.