Surgical Drainage of Symptomatic Peripancreatic Fluid Collections in the Era of Endoscopic Management
Luis a. Benavente-Chenhalls*1, Eduardo E. Montalvo-Jave3, Michael B. Farnell1, Michael G. Sarr1, Todd H. Baron2, Michael L. Kendrick1
1General Surgery, Mayo Clinic, Rochester, MN; 2Gastroenterology, Mayo Clinic, Rochester, MN; 3Gastroenterology Research Unit, Mayo Clinic, Rochester, MN
Background: With the advent of endoscopic drainage of symptomatic peripancreatic fluid collections (PPFC), the role of surgical intervention has decreased. Patients with complex, inaccessible collections or with prior unsuccessful drainage attempts account for an increasing proportion of those managed surgically. Our aim was to evaluate the morbidity, mortality and outcomes of the surgical management of symptomatic PPFC in this new era.
Methods: Retrospective review of all patients undergoing surgical management of benign, symptomatic PPFC from 1990 to 2006.
Results: Of 449 patients having undergone drainage procedures for PPFC, 105 (23%) had surgical drainage, comprising our study group. The mean age was 50 years (21-86), with 58% male. The mean interval from the first episode of pancreatitis to surgical intervention was 18 months. Surgical intervention was the primary treatment in 57%, or secondary after previous endoscopic (38%) or percutaneous (5%) attempts. The mean number of endoscopic procedures was 2.6 (1-14) and included transgastric (63%), transpapillary (27%), and transduodenal (7%) approaches. Operative intervention included cyst-enterostomy (47%), cyst-gastrostomy (31%), distal pancreatectomy (10%), external drainage (9%), lateral pancreaticojejunostomy (1%), and total pancreatectomy (1%). Evidence of peripancreatic necrosis at operation was evident in 8%. Major postoperative morbidity or perioperative mortality occurred in 41% and 0% respectively. Perioperative complications included infection-related complications (23%), anemia requiring transfusion (17%), pulmonary embolism (6%), pancreatic fistula (3%), and small bowel obstruction (3%). Perioperative ICU admission was necessary in 15%, with a mean ICU stay of 8 days (1-33). Perioperative reoperation for abdominal sepsis or hemorrhage was required in 4 patients. The mean hospital stay was 31 days (2-242). Recurrent pancreatitis or pseudocyst occurred in 17% and 9% respectively.
Conclusion: Surgical drainage of PPFC is less commonly performed in the era of endoscopic drainage. Operative intervention in these select patients is associated with a high morbidity and may be attributed to increased disease complexity preventing endoscopic drainage attempts, or to complicating factors after attempted drainage.