Initial experience with robotic liver resection: Audit of 120 consecutive cases at a single center and comparison with open and laparoscopic approaches

Background/Purpose Surgical outcomes and utility of robotic liver resection (RLR) are undefined. Methods We retrospectively studied perioperative and long-term outcomes of the single-center 120 RLRs including non-anatomic (NAR, n = 58) and anatomic (AR, n = 62) resections. To evaluate the feasibility and safety of RLR, perioperative outcomes of RLR (n = 103) were compared to those of open (OLR, n = 495) or laparoscopic (LLR, n = 451) resection in liver-only resections without reconstruction, using 1:1 propensity score matching (PSM). The changing trends from the earlier to the later RLR cases were assessed. Long-term outcomes were compared between RLR and LLR. Results Various types of RLR with different surgical difficulties were performed, with mostly comparable postoperative morbidity between AR and NAR, or among AR subtypes. In segmentectomy and sectionectomy cases, perioperative outcomes significantly improved in the later period. In comparison between PSM-selected OLR and RLR cases (87:87), RLR had significantly longer operative time, less blood loss, and shorter hospital stay. PSM-selected LLR and RLR cases (91:91) showed comparable perioperative outcomes. Overall and recurrence-free survivals after RLR for newly diagnosed hepatocellular carcinoma and colorectal metastasis were comparable to those after LLR. Conclusions RLR is applicable to various types of liver resection with acceptable perioperative and long-term outcomes in select patients.