Perioperative outcomes of laparoscopic pancreaticoduodenectomy for benign and borderline malignant periampullary disease compared to open pancreaticoduodenectomy

Purpose: The objective of this study was to compare perioperative outcomes between laparoscopic pancreaticoduodenectomy (LPD) and open pancreaticoduodenectomy (OPD) for benign and borderline malignant periampullary diseases. Methods: Of 107 pancreaticoduodenectomy cases for non-malignant diseases from March 1993 to July 2017, 76 patients underwent OPD and 31 patients received LPD. To adjust for baseline differences and selection bias, operative outcomes and complications were compared after propensity score matching (PSM). Results: After 1:1 PSM, well-matched 31 patients in each group were evaluated. As a result, significant differences were observed between two groups in some aspects: mean operative time (LPD 426.8 ± 98.58 vs. OPD 355.03 ± 100.0 min, p = 0.031), estimated blood loss (LPD 477.42 ± 374.80 vs. OPD 800.00 ± 531.35 ml, p = 0.008), and postoperative hospital stay (LPD 14.74 ± 5.40 vs. OPD 23.81 ± 11.63 days, p < 0.001). The average visual analogue scores for pain observed from patients in LPD group on postoperative day (POD) 1 (4.23 ± 1.83 vs. 5.55 ± 2.50, p = 0.021) and POD 3 (3.32 ± 1.66 vs. 5.26 ± 2.76, p = 0.002) were significantly less than those from patients in OPD group, as well. There were no significant differences between groups about major complications including the rate of postoperative pancreatic fistula. Conclusions: LPD is a safe procedure and provides less postoperative pain and the shortening length of hospitalization. LPD may serve the feasible alternative approach for benign and borderline malignant periampullary disease.