Comparative analysis of robot-assisted minimally invasive esophagectomy versus conventional minimally invasive esophagectomy, a systematic review and meta-analysis.

Robot-assisted minimally invasive esophagectomy (RAMIE) may enhance visualization and lymph-node dissection compared with conventional minimally invasive esophagectomy (cMIE), but comparative effectiveness remains uncertain. We performed a PRISMA-compliant systematic review and meta-analysis of randomized controlled trials and propensity-matched cohort studies comparing RAMIE with MIE (from 2013 to August 26th, 2025). PubMed, Cochrane Library, Embase, and Web of Science were searched. Continuous outcomes were pooled as mean differences (MD) and dichotomous outcomes as risk ratio (RR) using randomized-effects models; heterogeneity was assessed with I. Risk of bias was evaluated with RoB-2 for trials and the Newcastle-Ottawa Scale for cohorts. Twenty-five studies (~ 8,900 patients; RAMIE ≈ 4,200; MIE ≈ 4,700) met eligibility criteria. Total operative time was longer with RAMIE (MD 38.91 min, 95% CI 16.05-61.76; p = 0.0008; I²=97%), while thoracic operative time alone was not significantly different (MD 16.18 min, 95% CI - 2.46 to 34.82; p = 0.09; I²=94%). Estimated blood loss was modestly lower with RAMIE (MD - 12.73 mL, 95% CI - 25.25 to - 0.21; p = 0.05; I²=86%). RAMIE achieved a higher total lymph node yield (MD 2.01 nodes, 95% CI 1.05-2.96; p < 0.001; I²=57%) and retrieved more left RLN lymph nodes (MD 0.60, 95% CI 0.13-1.08; p = 0.01), with no significant difference in right RLN nodes (MD 0.11, 95% CI - 0.02 to 0.24; p = 0.09; I² = 31%). In-hospital mortality (RR 0.76, 95% CI 0.42-1.39; p = 0.38; I²=0%) and 90-day mortality (RR 0.97, 95% CI 0.53-1.77; p = 0.91; I²=0%) were similar. Overall complications (RR 0.90, 95% CI 0.79-1.02; p = 0.09) and major complications (RR 0.80, 95% CI 0.63-1.01; p = 0.06) did not differ. Pulmonary complications trended lower (RR 0.88, 95% CI 0.76-1.02; p = 0.10), and recurrent laryngeal nerve palsy was borderline lower (RR 0.79, 95% CI 0.62-1.00; p = 0.05). Length of stay was shorter with RAMIE (MD - 1.28 days, 95% CI - 2.26 to - 0.34; p = 0.01; I²=84%), whereas ICU stay was longer (MD 2.68 days, 95% CI 0.51-4.84; p = 0.02; I²=98%). Publication bias signals for thoracic operative time and length of stay were driven by outlier on sensitivity analysis. RAMIE and cMIE are both safe. RAMIE involves a longer operative time but yields lower estimated blood loss, a higher total lymph nodes harvest-particularly along the left RLN- fewer pulmonary complications by trend, and a shorter length of stay, with no significant difference in anastomotic leak or early mortality. These results support patient-centered selection between RAMIE and cMIE, taking center experience and the learning curve into account. PROSPERO CRD420251134020.