Management of peripancreatic fluid collections following partial pancreatectomy: a comparison of percutaneous versus EUS-guided drainage

Peripancreatic fluid collections (PFCs) occur in up to 30 % of patients following partial pancreatic resections. Traditionally, postoperative PFCs are managed via percutaneous drainage (PD). EUS-guided transgastric drainage has been shown to be effective for the management of PFCs secondary to acute pancreatitis. However, there are limited data on the use of EUS-guided drainage (EUSD) for the management of postoperative PFCs.

To compare the safety and efficacy of PD versus EUSD of PFCs in patients who have undergone partial pancreatic resections.

Retrospective analysis of patients with symptomatic PFCs following pancreatic enucleation or distal pancreatectomy at MSKCC between January 2008 and December 2010. Patients were identified using an electronic medical records query in addition to a prospectively maintained pancreatic surgery complications database.

Single, academic, tertiary referral center.

Twenty-three patients with symptomatic PFCs following pancreatic enucleation or distal pancreatectomy were retrospectively identified.

CT-guided PD versus endoscopic ultrasound-guided drainage (EUSD)

Outcomes included technical success, clinical success, number of interventions, and complications. Technical success was defined as successful localization of the PFC by fine-needle aspiration and placement of a drainage catheter or stent. Clinical success was defined as radiographic resolution of the PFC and removal of the drain or stent, without the need for an alternative drainage procedure or surgery.

PD was initially performed in 14 patients and EUSD initially in 9 patients. Three patients with initial PD had recurrence of PFC after removal of the external drain and underwent subsequent EUSD. The mean size of the fluid collections was 10.0 cm in the PD group and 8.9 cm in the EUSD group. Technical success was achieved in all patients in both groups. Clinical success was achieved in 11 of 14 (79 %) patients in the PD group compared with 11 of 11 (100 %) in the EUSD group, with one patient in the EUSD group lost to follow-up. One patient with initial PD required two additional percutaneous procedures before complete PFC resolution. Five EUSD patients required repeat endoscopy for stent revision or necrosectomy. The median number of interventions was two [range 1–5] in the PD group and two [range 1–5] in the EUSD group. The median number of days the drainage catheters were in place was 44.5 [range 2–87] in the PD group versus 57 [range 32–217] in the EUSD group. There were no procedure-related complications in either group. Adverse events in the PD group included splenic artery stump bleeding, pleural effusion, cysto-colonic and cysto-cutaneous fistulae, and persistent catheter site pain. One patient in the EUSD group developed a small-bowel obstruction and bleeding gastric ulcer.

Retrospective, nonrandomized study and small numbers.

EUSD of postoperative PFCs appears to be safe and technically feasible. This technique appears to be as successful as PD for the management of PFCs with the advantage of not requiring an external drainage apparatus and should be considered as a therapeutic option in this group of patients. Further evaluation, with analysis of cost and quality of life, should be considered in a prospective, randomized trial.