Establishing the Learning Curve of Laparoscopic and Robotic Distal Gastrectomy: a Systematic Review and Meta-Regression Analysis

Background Minimally invasive distal gastrectomy (MIDG) is non-inferior compared with open distal gastrectomy for gastric cancer. However, MIDG bears a learning curve (LC). This study aims to evaluate the number of cases required to surmount the LC (i.e. N-LC) in MIDG. Methods PubMed, Embase, Scopus, and the Cochrane Library were systematically searched from inception to August 2022 for studies which reported N-LC in MIDG. N-LC on reduced-port/single-port MIDG only were separately analysed. Poisson mean (95% confidence interval (CI)) was used to determine N-LC. Negative binomial regression was used to compare N-LC between laparoscopic distal gastrectomy (LDG) and robotic distal gastrectomy (RDG). Results A total of 45 articles with 71 data sets (LDG n=47, RDG n=24) were analysed. There were 7776 patients in total (LDG n=5516, RDG n=2260). Majority of studies were conducted in East Asia (n=68/71). Majority (76.1%) of data sets used non-arbitrary methods of analyses. The overall N-LC for RDG was significantly lower compared to LDG (RDG 22.4 (95% CI: 20.4-24.5); LDG 46.7 (95% CI: 44.1-49.4); incidence rate ratio 0.48, p<0.001). The median number of laparoscopic gastrectomy (LG) cases prior was 0 (interquartile range (IQR) 0-105) for LDG and 159 (IQR 101-305.3) for RDG. Meta-regression analysis did not show a significant impact prior experience in LG, extent of lymphadenectomy and intracorporeal vs extracorporeal anastomosis had on overall N-LC for LDG and RDG. Conclusion N-LC for RDG is shorter compared to LDG, but this may be due to prior experience in LG and ergonomic advantages of RDG.