Endoscopic treatment of postsurgical external pancreatic fistulas.

BACKGROUND AND STUDY AIMS

External pancreatic fistulas (EPFs) are managed primarily by conservative treatment with a success rate of 40-90%. Failures of conservative therapy have traditionally been dealt with using surgery; however, major morbidity and mortality are associated with operative treatment. The aim of this study was to evaluate the feasibility and effectiveness of endoscopic treatment in the closure of EPF.

PATIENTS AND METHODS

A total of 16 consecutive patients with EPF (12 men, four women; median age 50, range 21-66) underwent an attempt at endoscopic management after failure of conservative therapy. Four patients had chronic pancreatitis. All patients had EPFs occurring after open abdominal surgery. The mean interval between the onset of the fistula and our intervention was 108 days (range 27-365 days). The mean output volume of the fistula was 205 ml/d (range 50-600 ml/ d). The aim of treatment was to lower the pancreatic duct pressure and to bypass the ductal disruption by placement of drains and/or stents to induce fistula healing.

RESULTS

In all, 13 biliary and nine pancreatic sphincterotomies were performed in order to gain access to the pancreatic duct. Access through the minor papilla was required in one patient. Complete visualization of the main pancreatic duct as well as of the fistulous tract was obtained in 12 patients (75%). Treatment consisted of placement of a nasal pancreatic drain (NPD) across the pancreaticojejunal anastomosis in one patient after duodenopancreatectomy. In 11 of the remaining 15 patients (73%) a NPD could be placed in the pancreatic duct across the ductal leakage (n = 9) or nearby (n = 2). One patient died 24 hours after endoscopic treatment from severe sepsis and massive pulmonary embolism. Endoscopic drainage was effective in healing the EPF in all patients in whom NPDs had been successfully placed, except one. The fistula in this patient healed completely after insertion of an 8.5-Fr pancreatic stent. The mean interval between endoscopic treatment and fistula closure was 8.8 days (range 2-33 days). No complications related to the endoscopic treatment were recorded in this series. In the 12 successfully treated patients, fistulas did not recur in any of the 11 surviving patients after a mean follow-up of 24.7 months (range 3-63 months).

CONCLUSIONS

Endoscopic pancreatic drainage, when feasible, is safe and effective for EPF and should be considered as a first-line therapy when EPFs do not respond to conservative therapy.