Laparoscopic pancreatoduodenectomy with open or laparoscopic reconstruction during the learning curve: a multicenter propensity score matched study

Background: Laparoscopic pancreatoduodenectomy with open reconstruction (LPD-OR) has been suggested to lower the rate of postoperative pancreatic fistula reported after laparoscopic pancreatoduodenectomy with laparoscopic reconstruction (LPD). Propensity score matched studies are, lacking. Methods: This is a multicenter prospective cohort study including patients from 7 Dutch centers between 2014-2018. Patients undergoing LPD-OR were matched LPD patients in a 1:1 ratio based on propensity scores. Main outcomes were postoperative pancreatic fistulas (POPF) grade B/C and Clavien-Dindo grade >= 3 complications. Results: A total of 172 patients were included, involving the first procedure for all centers. All 56 patients after LPD-OR could be matched to a patient undergoing LPD. With LPD-OR, the unplanned conversion rate was 21% vs. 9% with LPD (P < 0.001). Median blood loss (300 vs. 400 mL, P = 0.85), operative time (401 vs. 378 min, P = 0.62) and hospital stay (10 vs. 12 days, P = 0.31) were comparable for LPD-OR vs. LPD, as were Clavien-Dindo grade >= 3 complications (38% vs. 52%, P = 0.13), POPF grade B/C (23% vs. 21%, P = 0.82), and 90-day mortality (4% vs. 4%, P > 0.99). Conclusion: In this propensity matched cohort performed early in the learning curve, no benefit was found for LPD-OR, as compared to LPD.