Feed

Lung transplantation is conventionally performed through clamshell thoracosternotomy or sternotomy, approaches associated with pain, impaired wound healing, and chronic neuralgia. Robotic assistance has reached transplantation only recently. We systematically reviewed the clinical experience with robotic-assisted lung transplantation. Following PRISMA 2020 and PRISMA-S, we searched MEDLINE, Embase, Scopus, Web of Science, and Cochrane CENTRAL from inception to 15 August 2026, with trial registries, conference abstracts, and Chinese-language sources, without language restriction. Human robotic lung transplant reports with at least one clinical outcome were eligible. Because only one comparative cohort existed, synthesis was narrative; no meta-analysis was performed. Certainty was graded by outcome. PROSPERO: CRD420261479535. Out of 393 database records and 7 from other methods, six reports from five reporting centers (27 unique recipients) were included. No randomized trial or prior full-length systematic review was identified. In the only comparative cohort (robotic n = 21 vs. direct-vision minimally invasive n = 90), one-year survival was 95.0% versus 95.5% (log-rank P = 0.84) and grade-3 primary graft dysfunction at 72 h 4.8% versus 8.9% (odds ratio 0.51, 95% CI 0.06-4.34), despite longer ischemic times (486 vs. 408 min; P = 0.02). Every robotic versus direct-vision confidence interval included unity. Certainty was very low for all outcomes. Robotic lung transplantation is technically feasible across five reporting centers. Comparative evidence rests on one clinically selected, non-randomized cohort in which no difference was statistically detectable; this establishes neither equivalence nor absence of harm. Multicenter prospective evaluation with standardized outcomes is required.Registration: The protocol was registered with PROSPERO (CRD420261479535).