Outcomes of endoscopic and percutaneous drainage of pancreatic fluid collections arising after pancreatic tail resection.

BACKGROUND: Up to 15% to 30% of patients develop pancreatic fluid collections (PFCs) after pancreatic tail resection. Percutaneous and endoscopic methods have been used to drain these collections, though few data are available that compare outcomes of these modalities.

STUDY DESIGN: From December 1998 to April 2011, we identified all patients who underwent pancreatic tail resection and developed PFCs requiring intervention. The primary aim was to compare overall success rates in resolution of PFCs using endoscopic and percutaneous modalities. Success rates, hospital length of stay, number of CT scans, sinograms and endoscopies performed, and days with drain(s) in place were compared.

RESULTS: Forty-eight patients were identified. Percutaneous drainage was performed a median of 25 days postoperatively, compared with 85 days for endoscopic drainage (p < 0.001). Endoscopic and percutaneous methods had similar rates of technical success (100% vs 97%, p = 0.50) and treatment success (80% vs 81%, p = 0.92), respectively. Recurrence rates were 16.6% for the endoscopic group and 23% for the percutaneous group (p = 0.65), and adverse events occurred in 9.4% of those treated endoscopically vs 13.3% of those treated percutaneously (p = 0.68). Location and characteristics of PFCs did not influence success rates. Recurrences were often treated by "salvage" drainage via the other modality. Median hospital stay was longer after primary percutaneous drainage compared with primary endoscopic drainage (5.5 days vs 2 days, p = 0.046). Primary percutaneous drainage patients also had more CT scans (median 3 vs 2, p = 0.03).

CONCLUSIONS: Endoscopic drainage and percutaneous drainage appear to be equally effective and complementary interventions for PFCs occurring after pancreatic tail resection. Primary endoscopic drainage may be associated with shorter hospital stay and fewer CT scans.