Short-term Outcomes of Robotic Versus Open Pancreatoduodenectomy
Felix Nickel, Philipp Anthony Wise, Philip C. Müller, Christoph Kuemmerli, Amila Cizmic, Gabriel Alexander Salg, Verena Steinle, Anna Nießen, Philipp Mayer, Arianeb Mehrabi, Martin Loos, Beat P. Müller‐Stich, Yakup Kulu, Markus W. Büchler, Thilo Hackert
Abstract
OBJECTIVE: The goal of the current study was to investigate the perioperative outcomes of robotic pancreaticoduodenectomy (RPD) and open pancreaticoduodenectomy (OPD) in a high-volume center. BACKGROUND: Despite RPDs prospective advantages over OPD, current evidence comparing the 2 has been limited and has prompted further investigation. The aim of this study was to compare both approaches while including the learning curve phase for RPD. METHODS: A 1:1 propensity score-matched analysis of a prospective database of RPD with OPD (2017-2022) at a high-volume center was performed. The main outcomes were overall- and pancreas-specific complications. RESULTS: Of 375 patients who underwent PD (OPD n=276; RPD n=99), 180 were included in propensity score-matched analysis (90 per group). RPD was associated with less blood loss [500 (300-800) vs 750 (400-1000) mL; P =0.006] and more patients without a complication (50% vs 19%; P <0.001). Operative time was longer [453 (408-529) vs 306 (247-362) min; P <0.001]; in patients with ductal adenocarcinoma, fewer lymph nodes were harvested [24 (18-27) vs 33 (27-39); P <0.001] with RPD versus OPD. There were no significant differences for major complications (38% vs 47%; P =0.291), reoperation rate (14% vs 10%; P =0.495), postoperative pancreatic fistula (21% vs 23%; P =0.858), and patients with the textbook outcome (62% vs 55%; P =0.452). CONCLUSIONS: Including the learning phase, RPD can be safely implemented in high-volume settings and shows potential for improved perioperative outcomes versus OPD. Pancreas-specific morbidity was unaffected by the robotic approach. Randomized trials with specifically trained pancreatic surgeons and expanded indications for the robotic approach are needed.