Back to Annual Meeting Posters

Introduction: Clinically significant pancreatic leak continues to complicate distal pancreatectomies (DP). We report the outcomes of various methods of pancreatic transection and management of the pancreatic stump at our institution. Methods: Retrospective review of all patients undergoing DP from 01/1999 to 07/2010. Leaks were retrospectively classified according to the strict ISGPF guidelines. Grade B and C leaks were grouped as clinically significant (CSL). Results: A total of 820 patients underwent DP, of which 147 (18 %) had a pancreatic leak. Leaks were classified as Grade A, B, and C in 57%, 42%, and 1% respectively. The median age at the time of surgery was 59 years (Range 15-90 years) and 53% were males. Clinical characteristics of patients with regard to age, sex, BMI, smoking status, benign or malignant disease, diabetic status, or blood transfusion did not differ significantly among the grades of pancreatic leak. Intra-operative administration of hetastarch was associated with CSL (p = 0.045). The pancreas was transected using the stapler in 51.9%, electrocautery 34.6%, ultrasonic scalpel 8.4%, saline coupled radio frequency ablation (RFA) 3.17% and scalpel 1.8%. The visible pancreatic duct and/or parenchyma were oversewn in 73%. In 21.6%, pancreatic stump was treated with the RFA device. Clinically significant leak was seen in 3.4% of patients whose pancreas was transected with a stapler and oversewn versus 15.3% of patients in whose pancreas was stapled. Patients whose pancreas was transected using the scalpel or an energy device and treated with RFA had a 13.3% CSL rate. Pancreas transected using a stapler and the stump treated with RFA had a 19.2% CSL rate, whereas oversewing a pancreatic margin that had been treated with the RFA device had a 28.6% clinically significant leak rate. A patient with transected margin treated with oversewn relative to a patient whose pancreas transected with stapler and oversewn was at highest risk for CSL [p = <0.001, OR 11.5 (CI 3.1 - 42.4)]. In univariate models, the use of the RFA device and oversewing of the pancreatic duct were predictors of a CSL (p<0.05). On evaluating various modes of transection, there was interaction of RFA with oversewing and stapling with oversewing of the pancreatic stump (p <0.001)]. Conclusion: Among various methods available for pancreatic transection during DP, many of them recent technologies, none have a clinical superiority. Using the stapler to transect the pancreas has a higher rate of clinically significant leak as compared to treating the transected stump with RFA. Using the RFA device in addition to a stapler or oversewing the transected margin has a higher rate of clinically significant leak and should not be attempted. Randomized trials of newer technologies to help solve this age old dilemma are necessary.