S2310 Comparing Robotic, Laparoscopic and Open Pancreaticoduodenectomy: A Systematic Review and Network Meta-Analysis of Randomized Controlled Trials and Propensity-Matched Studies

Introduction: Pancreaticoduodenectomy is a complicated procedure with a high risk of morbidity and mortality. This network meta-analysis compared the robotic (RPD), laparoscopic (LPD) and open (OPD) approaches for pancreaticoduodenectomy. Methods: PubMed, Cochrane Central and ScienceDirect were searched from inception till January 2025. A frequentist network meta-analysis was performed using the R version 4.2.1 and employing the “netmeta” package. Risk ratios (RR) with a 95% Confidence interval were pooled under the random effects model for the dichotomous outcomes. The treatment ranking was calculated using the P-scores. The quality of the included studies was evaluated using the Cochrane Rob 2 tool and the New-castle Ottawa Scale, while the publication bias was assessed using the funnel plots. Results: Fifty-five studies were included in the final meta-analysis. Compared with OPD, RPD significantly decreased the risk of overall complications (RR = 0.84, 95% CI: [0.72, 0.97). The risk of major complications was lowest in the RPD (P-score = 0.95) and highest in the OPD group (0.13). RPD also had the lowest risk of postoperative pancreatic fistula rate (POPF) (0.42), while the risk of delayed gastric emptying (DGE) was lowest in the LPD group (P-score = 1.0). The R0 resection rate was highest in the LPD (0.72) and lowest in the RPD group (0.39). Regarding 90-day mortality, RPD had the lowest risk (P-score = 0.72). Conclusion: RPD demonstrates a lower risk of overall complications and postoperative pancreatic fistula compared to OPD, while LPD has a lower risk of delayed gastric emptying and higher R0 resection rates. Both RPD and LPD offer promising outcomes, with RPD showing the most favourable results in reducing 90-day mortality.