Concomitant training in robotic and laparoscopic liver resections of low-to-intermediate difficulty score: a retrospective analysis of the learning curve.

In the setting of minimally invasive liver surgery (MILS), training in robotic liver resections (RLR) usually follows previous experience in laparoscopic liver resections (LLR). The aim of our study was to assess the learning curve of RLR in case of concomitant training with LLR. We analyzed consecutive RLRs and LLRs by a surgeon trained simultaneously in both techniques (Surg1); while a second surgeon trained only in LLRs was used as control (Surg2). A regression model was used to adjust for confounders and a Cumulative Sum (CUSUM) analysis was carried out to assess the learning phases according to operative time and difficulty of the procedures (IWATE score). Two-hundred-forty-five procedures were identified (RobSurg1, n = 75, LapSurg1, n = 102, LapSurg2, n = 68). Mean IWATE was 4.0, 4.3 and 5.8 (p < 0.001) in each group. The CUSUM analysis of the adjusted operative times estimated the learning phase in 40 cases (RobSurg1), 40 cases (LapSurg1), 48 cases (LapSurg2); for IWATE score it was 38 cases (RobSurg1), 33 cases (LapSurg1), 38 cases (LapSurg2) respectively. Our preliminary experience showed a similar learning curve of 40 cases for low and intermediate difficulty RLR and LLR. Concomitant training in both techniques was safe and may be a practical option for starting a MILS program.

Robotic versus laparoscopic liver resections: propensity-matched comparison of two-center experience.

BACKGROUND: The advantages of the robotic approach in minimally invasive liver surgery (MILS) are still debated. This study compares the short-term outcomes between laparoscopic (LLR) and robotic (RLR) liver resections in propensity score matched cohorts. METHODS: Data regarding minimally invasive liver resections in two liver surgery units were retrospectively reviewed. A propensity score matched analysis (1:1 ratio) identified two groups of patients with similar characteristics. Intra- and post-operative outcomes were then compared. The difficulty of MILS was based on the IWATE criteria. RESULTS: Two hundred sixty-nine patients underwent MILS between January 2014 and December 2021 (LLR = 192; RLR = 77). Propensity score matching identified 148 cases (LLR = 74; RLR = 74) consisting of compensated cirrhotic patients (100%) underwent non-anatomic resection of IWATE 1-2 class (90.5%) for a solitary tumor < 5 cm in diameter (93.2%). In such patients, RLRs had shorter operative time (227 vs. 250 min, p = 0.002), shorter Pringle's cumulative time (12 vs. 28 min, p < 0.0001), and less blood loss (137 vs. 209 cc, p = 0.006) vs. LLRs. Conversion rate was nihil (both groups). In RLRs compared to LLRs, R0 rate (93 vs. 96%, p > 0.71) and major morbidity (4.1 vs. 5.4%, p > 0.999) were similar, without post-operative mortality. Hospital stay was shorter in the robotic group (6.2 vs. 6.6, p = 0.0001). CONCLUSION: This study supports the non-inferiority of RLR over LLR. In compensated cirrhotic patients underwent resection of low-to-intermediate difficulty for a solitary nodule < 5 cm, RLR was faster, with less blood loss despite the shorter hilar clamping, and required shorter hospitalization compared to LLR.