Back to Annual Meeting Program

Background: POEM is a promising totally endoscopic method for treating achalasia. Esophageal perforation and incomplete myotomy are technical errors with severe adverse consequences, hence appropriate training is essential for the safe adoption of this procedure. The objective of this study was to compare swine and fresh human cadavers(FHC) as training models for POEM.
Methods: Healthy Yorkshire male pigs(40-46 kg) were used for POEM training. Procedures were also performed in male and female FHC during the same time period. A standardized procedure was used in both models. Following submucosal injection of methylene blue dye to mark the distal extent of the dissection, a mucosotomy was made in the mid-esohagus. The endoscope was then inserted into the mucosotomy and a submucosal tunnel bluntly dissected. Endoscopic myotomy of the circular muscle layer from the mid-esophagus to the gastroesophageal junction was performed using a triangular tip knife. The mucosotomy was subsequently closed with endoscopic clips. Following POEM in swine, the animals were sacrificed and necropsy performed to assess for organ injury and myotomy adequacy. In FHC, thoracoscopic and laparoscopic evaluation was performed to assess for pleural violation,organ injury and myotomy adequacy. Adequacy of myotomy was determined by endoscopic transillumination at distal most myotomy site.
Results: POEM was performed in 7 acute swine and 6 FHC. In swine, POEM was successfully completed in 5 animals(72%). Two animals(29%) expired during the procedure from cardiopulmonary collapse. In the remaining 5 animals, POEM was uncomplicated with no evidence of complication on necropsy. The average procedure time was 90minutes(range70-120minutes). The primary limitation of the swine model related to the attenuated circular muscle of the porcine esophagus which made myotomy technically difficult and pleural violation a frequent complication. No problems closing the myotomy were encountered. In contrast, POEM was successfully performed in all FHC. The average procedure time was 85minutes(70-120minutes). Pleural violation was noted in one cadaver. In 4 cadavers, the mucosotomy was inadvertently extended during creation of the submucosal tunnel. The technical limitations of the cadaver model were difficulties with the mucosotomy and submucosal tunnel dissection due to decreased tissue pliability and poor tissue distension.
Conclusion: Although cadavers are more anatomically relevant as a training model for POEM, creation of the mucosotomy and submucosal tunnel is limited by poor tissue pliability. Conversely, mucosotomy and submucosal tunnel dissection is more easily achieved in an acute swine model, but the myotomy limited by attenuated circular muscle. A hybrid teaching paradigm using both porcine and human cadaver model may be necessary for comprehensive POEM training.