Systematic review and meta-analysis of robotic versus open pancreaticoduodenectomy

Background Although robotic pancreaticoduodenectomy (RPD) has been successfully performed since 2003, its advantages over open pancreaticoduodenectomy (OPD) are still uncertain. The aim of this systematic review and metaanalysis was to compare the clinical outcomes of RPD to those of OPD. Methods A systematic literature review was performed to identify RPD versus OPD comparative studies published between January 2003 and January 2016. Intraoperative outcomes, post-operative outcomes and oncologic safety were evaluated. Pooled odds ratios (ORs) and weighted mean differences (WMDs) with a 95% confidence interval (95% CI) were calculated using fixed-effect or randomeffect models. Results Nine non-randomized observational clinical studies involving 680 patients met the inclusion criteria and involved 245 RPDs and 435 OPDs. The overall complication rate was significantly lower in RPD (OR 0.65, 95% CI 0.47-0.91, P = 0.012), as well as the margin positivity rate (OR 0.40, 95% CI 0.20-0.77, P = 0.006), the wound infection rate (OR 0.18, 95% CI 0.06-0.53, P = 0.002) and the length of hospital stay (WMD = -6.00, 95% CI -9.80 to -2.21, P = 0.002). There was no significant difference in the following: the number of lymph nodes harvested; the operation time; the reoperation rate; the incidence of delayed gastric emptying, bile leakage, pancreatic fistula and clinically significant pancreatic fistula; and mortality. The mean conversion rate was 7.3% (range 0-14%). Conclusions According to the results of this meta-analysis, RPD is as safe and efficient as OPD and is even favourable in terms of margin-negative resection, overall complication and wound infection rates and length of hospital stay. Given that there have not yet been any high-quality randomized controlled trials (RCTs), the evidence is still limited. Additional prospective, multi-centre RCTs are needed to further define the role of the robotic technique in PD.