Peroral endoscopic esophageal myotomy: defining the learning curve

Background: Peroral endoscopic myotomy (POEM) is an endoscopic alternative to laparoscopic esophageal myotomy. It requires a demanding skill set that involves both advanced endoscopic skills and knowledge of surgical anatomy and complication management. Objective: Determine the learning curve for POEM. Design: Prospective cohort study. Setting: Tertiary-care teaching hospital. Patients: The study involved the first 40 consecutive patients undergoing the POEM procedure under a prospective institutional review board protocol (research. gov #NCT01399476, 1056). Intervention: Peroral endoscopic myotomy for esophageal motility disorders. Main Outcome Measurements: Length of procedure (LOP) and technical errors (inadvertent mucosotomy). Results: A total of 40 patients underwent POEM. The mean LOP was 126 +/- 41 minutes. The mean myotomy length was 9 cm (range, 6-20 cm). The LOP per centimeter myotomy and variability decreased as our experience progressed. The means (+/- standard deviation) of the LOP per centimeter myotomy were as follows: first cohort, 16 +/- 4 minutes; second, 17 +/- 5 minutes; third, 13 +/- 3 minutes; fourth, 15 +/- 2 minutes; and fifth, 13 +/- 4 minutes. The incidence of inadvertent mucosotomy also decreased with increasing experience, to 8, 6, 4, 0, and 1, respectively. These minor complications were repaired intraoperatively with clips. There were 7 patients with capnoperitonium and another with bilateral capnothoraces that were associated with hemodynamic instability but resolved by Veress needle decompression. Two patients required endoscopy in the early postoperative period: self-limited hematemesis in one and radiologic evidence of leakage at the mucosotomy site in another. Limitations: Nonrandomized study. Conclusion: Mastery of operative technique in POEM is evidenced by a decrease in LOP, variability of minutes per centimeter of myotomy, and incidence of inadvertent mucosotomies and plateaus in about 20 cases for experienced endoscopists. The learning curve can be shortened with very close supervision and/or proctoring. (Gastrointest Endosc 2013; 77: 719-25.)