A Propensity Score-Matched Comparison of Robotic Versus Laparoscopic Gastrectomy for Gastric Cancer: Oncological, Cost, and Surgical Stress Analysis.

BACKGROUND: Robotic-assisted gastrectomy (RAG) has been rapidly adopted for gastric cancer (GC) treatment. However, whether RAG provides any significant outcome/cost advantages over laparoscopy-assisted gastrectomy (LAG) for the experienced laparoscopist remains unclear.
METHODS: A retrospective review of a prospectively collected database identified 768 consecutive patients who underwent either RAG (n = 103) or LAG (n = 667) for GC between July 2016 and June 2017 at a large center. A 1:3 matched propensity score analysis was performed. The short-term outcomes and hospital costs between the two groups were compared.
RESULTS: A well-balanced cohort of 404 patients was analyzed (RAG:LAG = 1:3 match). The mean operation times were 226.6 ± 36.2 min for the RAG group and 181.8 ± 49.8 min for the LAG group (p < 0.001). The total numbers of retrieved lymph nodes were similar in the RAG and LAG groups (means 38 and 40, respectively, p = 0.115). The overall and major complication rates (RAG, 13.9% vs. LAG, 12.5%, p = 0.732 and RAG, 3.0% vs. LAG, 1.3%, p = 0.373, respectively) were similar. RAG was much more costly than LAG (1.3 times, p < 0.001) mainly due to the amortization and consumables of the robotic system. According to cumulative sum (CUSUM), the learning phases were divided as follows: phase 1 (cases 1-21), phase 2 (cases 22-63), and phase 3 (cases 64-101), in the robotic group. The surgical stress (SS) was higher in the robotic group compared with the laparoscopic group in phase 1 (p < 0.05). However, the SS did not differ significantly between the two groups in phase 3.
CONCLUSIONS: RAG is a feasible and safe surgical procedure for GC, especially in the post-learning curve period. However, further studies are warranted to evaluate the long-term oncological outcomes and to elucidate whether RAG is cost-effective when compared to LAG.