Learning Curve of Robot-Assisted Lymph Node Dissection of the Left Recurrent Laryngeal Nerve: A Retrospective Study of 417 Patients.

ABSTRACT

OBJECTIVE: Left recurrent laryngeal nerve (no.106recL) lymph node dissection is a challenging procedure, and robotic-assisted minimally invasive esophagectomy (RAMIE) may have some advantages. This study aimed to determine the learning curve of no.106recL lymph node dissection. METHODS: The data of 417 patients who underwent McKeown RAMIE between June 2017 and June 2022 were retrospectively analyzed. The lymph node harvest of no.106recL was used to determine the learning curve, and the cumulative sum (CUSUM) method was employed to obtain the inflection point. RESULTS: A total of 404 patients (404/417, 96.9%) underwent robotic surgery. Based on the number of no.106recL lymph nodes harvested, the CUSUM learning curve was mapped and divided into three phases phase I (1‒75 cases), phase II (76‒240 cases), and phase III (241‒404 cases). The median (IQR) number of no.106recL lymph node harvests were 1 (4), 3 (6,) and 4 (4) in each phase (p < 0.001). The lymph node dissection rate gradually increased from 62.7% in phase I to 82.9% in phase III (p = 0.001). The total and thoracic lymph node harvest gradually increased (p < 0.001), whereas operation time (p = 0.001) and blood loss gradually decreased (p < 0.001). Moreover, the incidence of total complication (p = 0.020) and recurrent laryngeal nerve injury (p = 0.001) significantly decreased, and the postoperative hospital stay gradually shortened (p < 0.001). CONCLUSION: Robotic no.106recL lymph node dissection has some advantages for patients with esophageal cancer. In this study, perioperative and clinical outcomes were significantly improved over the learning curve. However, further prospective studies are required to confirm our results.