Short- and Long-Term Outcomes of Robot-Assisted Versus Video-Assisted Thoracoscopic Esophagectomy for Esophageal and Esophagogastric Junction Cancer

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Background/Objectives: Thoracoscopic esophagectomy is an accepted minimally invasive approach for resectable esophageal and esophagogastric junction cancer in Japan. Whether robot-assisted minimally invasive esophagectomy (RAMIE) improves short- and long-term outcomes compared with video-assisted thoracoscopic esophagectomy (VATS) remains clinically relevant. Methods: We retrospectively reviewed patients who underwent VATS or RAMIE between January 2017 and April 2026. Patients who underwent open esophagectomy, non-gastric conduit reconstruction, or staged operation were excluded. Propensity score matching (PSM) was performed using age, sex, body mass index, histology, tumor location, clinical stage, preoperative therapy, extent of lymph node dissection, abdominal approach, and reconstruction route. Long-term outcomes were explored in a PSM cohort limited to patients treated between January 2017 and December 2023. Results: The main cohort included 369 patients (VATS, n = 186; RAMIE, n = 183). Overall complications of Clavien–Dindo (CD) grade ≥ II were comparable between VATS and RAMIE (32.8% vs. 32.8%, p = 0.999), whereas RLNP CD ≥ I (9.7% vs. 3.8%, p = 0.025) and RLNP CD ≥ II (3.8% vs. 0.5%, p = 0.034) were less frequent after RAMIE. PSM yielded 128 matched pairs; RLNP remained less frequent after RAMIE (CD ≥ I, 10.9% vs. 3.1%, p = 0.015; CD ≥ II, 4.7% vs. 0%, p = 0.013), while other short-term outcomes were comparable. In the survival PSM cohort (79 pairs), 3-year overall survival (66.0% vs. 73.4%, p = 0.203) and recurrence-free survival (62.3% vs. 59.3%, p = 0.651) did not differ significantly between VATS and RAMIE. Conclusions: RAMIE showed comparable short-term safety to VATS and was associated with lower RLNP after adjustment for clinicopathological, treatment-related, and surgical factors. Exploratory long-term analysis was limited by cohort size and follow-up duration but suggested that RAMIE did not compromise oncologic outcomes.

This publication has 26 references indexed in Scilit: