Positive Surgical Margins After Anterior Robot-assisted Radical Prostatectomy: Assessing the Learning Curve in a Multi-institutional Collaboration

Background: The learning curve for robot-assisted radical prostatectomy (RARP) remains controversial, with prior studies showing that, in contrast with evidence on open and laparoscopic radical prostatectomy, biochemical recurrence rates of experienced versus inexperienced surgeons did not differ. Objective: To characterize the learning curve for positive surgical margins (PSMs) after RARP. Design, setting, and participants: We analyzed the data of 13 090 patients with prostate cancer undergoing RARP by one of 74 surgeons from ten institutions in Europe and North America between 2003 and 2022. Outcome measurements and statistical analysis: Multivariable models were used to assess the association between surgeon experience at the time of each patient's operation and PSMs after surgery, with adjustment for preoperative prostate-specific antigen level, grade, stage, and year of surgery. Surgeon experience was coded as the number of robotic radical prostatectomies done by the surgeon before the index patient's operation. Results and limitations: Overall, 2838 (22%) men had PSMs on final pathology. After adjusting for case mix, we found a significant, nonlinear association between surgical experience and probability of PSMs after surgery, with a lower risk of PSMs for greater surgeon experience (p < 0.0001). The probabilities of PSMs for a patient treated by a surgeon with ten, 250, 500, and 2000 prior robotic procedures were 26%, 21%, 18%, and 14%, respectively (absolute risk difference between ten and 2000 procedures: 11%; 95% confidence interval: 9%, 14%). Similar results were found after stratifying patients according to extracapsular extension at final pathology. Results were also unaltered after excluding surgeons who had moved between institutions. Conclusions: While we characterized the learning curve for PSMs after RARP, the relative contribution of surgical learning to the achievement of optimal outcomes remains controversial. Future investigations should focus on what experienced surgeons do to avoid positive margins and should explore the relationship between learning, margin rate, and biochemical recurrence. Understanding what margins affect recurrence and whether these margins are trainable or a result of other factors may shed light on where to focus future efforts in surgical education. Patient summary: In patients receiving robotic radical prostatectomy for prostate cancer, we characterized the learning curve for positive margins. The risk of surgical margins decreased progressively with increasing experience, and plateaued around the 500th procedure. Understanding what margins affect recurrence and whether these margins are trainable or a result of other factors has implications for surgeons and patients, and it may shed light on where to focus future efforts in surgical education.