INTRODUCTION
Postoperative pancreatic fistula (POPF) is one of the most prevalent clinically relevant complications following partial pancreatic resection. Endoscopic approaches have proven successful, but the literature regarding the best route of drainage is scarce. Our study was aimed at comparing the efficacy and safety of transpapillary (endoscopic retrograde cholangiopancreatography [ERCP]-based) and transmural (endoscopic ultrasound [EUS]-guided) endoscopic treatment of POPF occurring after distal pancreatectomy.
METHODS
Our observational and analytical retrospective cohort study included all patients who had undergone distal pancreatectomy in our institution since 2000 (173 patients). The primary endpoint was the clinical success of endoscopic treatment, defined as a complete resolution of the fistula and/or pancreatic fluid collection (or size decrease to < 2cm), with symptoms resolution, and without the need for percutaneous drainage or surgery. Secondary endpoints included the technical success (feasibility and efficacy of stent placement), the complication rate of endoscopic procedures, and the reintervention rate. Categorical variables were compared using the χ² test. Normally distributed continuous variables were analyzed by the Student t-test, and non-normally distributed variables by the Mann-Whitney U-test. Patients were divided into three groups (ERCP only, EUS drainage only, both EUS and ERCP).
RESULTS
Out of 173 surgical patients, 58 (33.5%) developed grade B and C POPF and were treated by endoscopy. The fistula rate was correlated neither with the surgery type (+-splenectomy, body and tail resection or solely tail, enucleation, isthmectomy, combined with other organ resections), neither with the indication for surgery (malignancy, trauma, pancreatitis). Fistulas were more severe in older patients (p=0.043). Patients were treated by ERCP alone (n=31, 53.4%), EUS alone (n=13, 22.4%), or both procedures (n=14, 24.1%). A significant shift was observed from ERCP alone (100% of cases in 2000-2005) towards EUS alone and combined EUS-ERCP (23% in 2006-2010, and 48% in 2011-2016). Technical success rates were similar in all groups (87-100-88%). In patients treated by ERCP only, the clinical success was 64.5% (20/31), as against 96.3% (26/27) in patients in whom EUS was performed at any points during endoscopic treatment (p=0.003). The overall reintervention rate was 44.8%, with a significantly lower reintervention rate when EUS was part of the treatments (20 and 23% vs. 55%, p<0.05). The ERCP- and EUS-related complication rates were 9.5% (7/75) and 17% (6/35), respectively (p=0.237), with only two complications recorded over the last 5 years.
CONCLUSIONS
Endoscopic treatment was highly successful in managing POPF following distal pancreatectomy, with a significantly higher clinical success rate of EUS compared to ERCP, resulting in less reinterventions. We therefore suggest considering EUS as a primary approach, while reserving ERCP for cases with pancreatic ductal strictures or inaccessible post-operative collections or fistulas.