Stepwise Improvement of Surgical Quality in Robotic Lateral Pelvic Node Dissection: Lessons From 100 Consecutive Patients With Locally Advanced Rectal Cancer

摘要

BACKGROUND: Lateral pelvic node dissection has significant technical difficulty and a high incidence of surgical morbidity. A steep learning curve is anticipated in performing lateral pelvic node dissection. However, no study has previously analyzed the learning curve and surgical skill acquisition for this complex procedure. OBJECTIVES: We aimed to evaluate the learning process for performing robotic total mesorectal excision with lateral pelvic node dissection in patients with rectal cancer. DESIGN: This is a retrospective analysis of a prospectively collected database. SETTING: This study was conducted at a tertiary cancer center. PATIENTS: A total of 100 patients who underwent robotic total mesorectal excision with lateral pelvic node dissection between 2011 and 2017 were included. MAIN OUTCOME MEASURES: A cumulative sum analysis was calculated based on the number of unilateral retrieved lateral pelvic nodes. Operative time, estimated bloodloss, lateral pelvic node metastatic rate, postoperative morbidities, and local recurrence were also analyzed. RESULTS: Cumulative sum modeling suggested 4 learning phases: learning I (33 patients), learning II (19 patients), consolidation (30 patients), and competence (18 patients). In the consolidation and competence phases, we adopted fluorescence imaging and standardized the surgical procedure on the basis of anatomical planes. The competence phase had the greatest number of unilateral retrieved lateral pelvic nodes (12.8 vs 4.9, 8.2, and 10.4; p < 0.001). Urinary complications, including urinary retention and postoperative α-blocker usage, were more frequently observed in learning phase I than in the competence phase (39.4% vs 16.7%, p = 0.034). During the median follow-up of 44.2 months, local recurrence in the pelvic sidewall was observed in 4 patients from learning phase I and in 1 patient from learning phase II. LIMITATIONS: This study was limited by its retrospective design. CONCLUSION: Completeness of the lateral pelvic node dissection procedure increased with the surgeon's experience and as new imaging systems and surgical technique standardization were implemented. Further studies are warranted to determine the oncologic outcomes associated with each phase. See Video Abstract at http://links.lww.com/DCR/B774. MEJORA GRADUAL DE LA CALIDAD QUIRRGICA EN LA DISECCIN LINFTICA PLVICA LATERAL ROBTICA LECCIONES DE PACIENTES CONSECUTIVOS CON CNCER DE RECTO LOCALMENTE AVANZADO: ANTECEDENTES:La disección linfática pélvica lateral tiene una dificultad técnica significativa y una alta incidencia de morbilidad quirúrgica. Se prevé una curva de aprendizaje muy pronunciada al realizar la disección linfática pélvica lateral. Sin embargo, ningún estudio ha analizado previamente la curva de aprendizaje y la adquisición de habilidades quirúrgicas para este procedimiento.OBJETIVOS:Nuestro objetivo fue evaluar el proceso de aprendizaje para realizar la escisión total de mesorrecto robótica con disección linfática pélvica lateral en pacientes con cáncer de recto.DISEÑO:Este es un análisis retrospectivo de una base de datos recopilada prospectivamente.AJUSTE:Este estudio se realizó en un centro oncológico terciario.PACIENTES:Un total de 100 pacientes fueron sometidos a escisión total de mesorrecto robótica con disección linfática pélvica lateral entre 2011 y 2017.PRINCIPALES MEDIDAS DE DESENLACE:Se calculó un análisis de suma acumulativa basado en el número unilateral de ganglios pélvicos laterales recuperados. También se analizaron el tiempo operatorio, la pérdida de sangre estimada, la tasa de metástasis ganglionares pélvicas laterales, las morbilidades postoperatorias y la recidiva local.RESULTADOS:El modelado total acumulativo sugirió cuatro fases de aprendizaje: aprendizaje I (33 pacientes), aprendizaje II (19 pacientes), consolidación (30 pacientes) y competencia (18 pacientes). En las fases de consolidación y competencia, adoptamos imágenes de fluorescencia y estandarizamos el procedimiento