The outcome of Laparoscopic pancreaticoduodenectomy is better with the learning curve. analysis in 100 patients

Abstract

Introduction: In our first experience Laparoscopic pancreaticoduodenectmoy (LPD) was associated with higher morbidity and we restricted our indications to patients at lower risk of pancreatic fistula (PF). Our aim was to analyse the learning curve effect on the outcome. Method: Between April 2011 – September 2017, 100 LPD were performed by pure laparoscopic approach and one layer pancreaticojejunal anastomosis. The demographics, surgical and the outcome data of the last 50 LPD were compared to the first 50 LPD. Results: No difference regarding age (57 vs 60, p = 0.20), female (42% vs 42%) and BMI (24 vs 23, p = 0.39). Less resections for pancreatic adenocarcinoma (22% vs 32%, p = 0.047) and for ampulloma (14 vs 26, p = 0.26). No difference on the operative time (335 vs 342, p = 0.62) and conversion (8% vs 8%) but Less blood loss (250 vs 368 vs, p = 0.024) and transfusion (4% vs 12%, p = 0.50). Similar 90 days mortality (4% vs 4%). Less overall morbidity (68% vs 76%, p = 0.37), less grades B and C PF (33% vs 42%, p = 0.33), less bleeding (18% vs 26%, p = 0.47), less re-interventions for severe complications (6 % vs 14%, p = 0.039), less readmission (4% vs 8%, p = 0.40), less delayed gastric emptying (10% vs 16%, p = 0.37), similar drained collections (4% vs 4%) but more biliary fistula (18% vs 8%, p = 0.137). Hospital stay was shorter (22 vs 28, p = 0.033). Conclusions: Although in the second period less resection were done for pancreatic adenocarcinoma, at lower risk of PF, the outcome of LPD was better related to patient selection, improved surgical technique and management.