Background: Robotic assistance enhances surgical dexterity and could facilitate wider adoption of laparoscopy for pancreatic resections (PR).
Methods: Data were prospectively entered into a database and analyzed retrospectively to assess feasibility and safety of robotic-assisted PR (RAPR). Additionally, robotic-assisted pancreaticoduodenectomy (RAPD) was compared to a contemporary group of open pancreaticoduodenectomies (OPD).
Results: Between October 2008 and October 2014, 200 consecutive patients underwent RAPR. Three procedures were converted to open surgery (1.5 %), despite 14 patients required associated vascular procedures. RAPD was performed in 83 patients (41.5 %), distal pancreatectomy in 83 (41.5 %), total pancreatectomy in 17 (8.5 %), tumor enucleation in 12 (6 %), and central pancreatectomy in 5 (2.5 %). Thirty-day and 90-day mortality rates were 0.5 and 1 %, respectively. Both deaths occurred after RAPD with vein resection. Complications occurred in 63.0 % of the patients (≥Clavien-Dindo grade IIIb in 4 %). Median comprehensive complication index was 20.9 (0-26.2). Incidence of grade B/C pancreatic fistula was 28.0 %. Reoperation was required in 14 patients (7.0 %). The risk of reoperation decreased after post-operative day 20 (OR 0.072) (p = 0.0015). When compared to OPD, RAPD was associated with longer mean operative time (527.2 ± 166.1 vs. 425.3 ± 92.7; <0.0001) but had an equivalent safety profile. The median number of examined lymph nodes (37; 28.8-45.3 vs. 36; 28-52.8) and the rate of margin positivity in patients diagnosed with pancreatic cancer were also similar (12.5 vs. 45.5 %).
Conclusions: RAPR, including RAPD, are safely feasible in selected patients. The results of RAPD in pancreatic cancer are encouraging but deserve further investigation.