Learning curve for lateral lymph node dissection in rectal cancer - a systematic review of literature.

BACKGROUND: Lateral lymph node dissection (LLND) remains controversial owing to differences in oncological principles between East and West, complex pelvic anatomy and the risk of complications. The aim of this systematic review is to determine the number of cases required to achieve surgical competence in LLND and to evaluate postoperative outcomes across different phases of the learning curve. METHODS: A systematic literature search was conducted in PubMed and Google Scholar for studies analyzing the LLND learning curve in rectal cancer resection. The three-phase pattern, consisting of learning, competence, and proficiency, was followed for data analysis and presentation. A separate learning curve analysis for open, laparoscopic and robotic LLND was performed. Blood loss, operative time, lymph node yield, urinary complications and postoperative morbidity were assessed across the phases of the learning curve for robotic LLND. RESULTS: Of the 616 articles screened, eight studies met the inclusion criteria. Seven studies reported the learning curve analysis for robotic LLND, and one study for laparoscopic and open approach. Five studies had operative time as a learning outcome, two studies the lymph node yield and one study both lymph node yield and urinary retention. All studies used the cumulative sum (CUSUM) method for learning curve analysis. Regardless of learning outcome, surgical competence for robotic LLND was achieved after 12-53 cases, for laparoscopic LLND after 19 cases, and for the open approach no inflection point was identified. In robotic LLND, blood loss, urinary complications, and morbidity decreased during the proficiency phase. CONCLUSIONS: The LLND learning curve is not yet standardized owing to variability in study design, type of LLND, and learning outcomes. Further well-designed and methodologically consistent studies are required to establish learning benchmarks and improve patient outcomes. REGISTRATION IN PROSPERO DATABASE: CRD420251050015.