Surgical outcomes of 500 robot-assisted minimally invasive esophagectomies for esophageal carcinoma

摘要

Background: In 2003, robot-assisted minimally invasive esophagectomy (RAMIE) was first reported to overcome the technical limitations of minimally invasive esophagectomy. RAMIE requires repeated modifications to set up the robotic system, and sufficient experience is required to gain technical proficiency. This study aimed to identify the learning periods and the outcomes of RAMIE for esophageal carcinoma. Methods: We retrospectively reviewed 500 consecutive RAMIE cases for esophageal cancer from December 2008 to February 2021. The learning curve for RAMIE was identified using cumulative sum analysis. Results: In a total of 500 RAMIE patients, the Ivor Lewis and McKeown operation were performed in 267 patients (53.4%) and 192 patients (38.4%), respectively. We classified learning periods into the learning phase (first 50 cases), the developing phase (51–150 case), and the stable phase (151–500 case). The rates of vocal cord palsy (42.0% vs. 28.4%) and anastomotic leakage (10.0% vs. 6.4%) were reduced after the learning phase. The mean total operative time (420 vs. 373 min), the mean length of stay (21.6 vs. 16.7 days), and the rate of anastomotic stricture (27.0% vs. 12.4%) were significantly reduced after reaching stable phase. In the stable phase, the proportion of the Ivor Lewis operation (26.0% vs. 67.1%), neoadjuvant chemoradiation therapy (14.0% vs. 25.7%), and bilateral cervical node dissection cases (12.0% vs. 22.0%) were significantly increased. Conclusions: Fifty procedures might be needed to achieve early proficiency, and extensive experience of more than 150 procedures is needed for quality stabilization.