Propensity score-matched analysis and mixed-effects logistic regression analysis comparing robotic versus laparoscopic liver resections in patients with cirrhosis complicated by portal hypertension: An international multicenter study

Background: Minimally invasive liver resection is increasingly performed, even in patients with cirrhosis and clinically significant portal hypertension. However, the benefits of robotic liver resection over laparoscopic liver resection in patients with cirrhosis and clinically significant portal hypertension remain unclear. Methods: We conducted a post hoc international, multicenter study including 2,328 patients with cirrhosis and clinically significant portal hypertension who underwent minimally invasive liver resection (2,004 laparoscopic liver resection and 324 robotic liver resection). Propensity score matching was performed in 1:1 and 2:1 ratios to compare perioperative outcomes between the 2 approaches. A mixed-effects logistic regression model with center as a random effect was additionally applied to the full unmatched cohort as a complementary analysis to account for center-level clustering. Results: After 1:1 propensity score matching (n = 287 per group), robotic liver resection was associated with significantly shorter operative time (180 vs 210 minutes, P = .002), reduced estimated blood loss (100 vs 200 mL, P < .001), lower open conversion rate (3.8% vs 9.2%, P = .012), lower postoperative morbidity rate (15% vs 26.5%, P = .002), and shorter postoperative stay (5 vs 6 days, P < .001). However, mixed-effects logistic regression on the entire unmatched cohort demonstrated that robotic liver resection was only associated with a significantly lower postoperative morbidity at the expense of a longer operation time. Conclusion: Both robotic liver resection and laparoscopic liver resection can be performed safely in selected patients with cirrhosis and clinically significant portal hypertension. Robotic liver resection was associated with lower postoperative morbidity.