The ESGE guideline, published in 2019, does not recommend performing the procedure directly on humans. It recommends performing procedures on at least 20 animal models before starting ESD procedures on humans, and performing en bloc resection without perforation in eight of the last ten procedures. Most animal models used today use pig organs due to their similarity to human organs. However, due to legal restrictions on pig production and slaughter in our country, access to pig models is difficult compared to other countries. Therefore, an animal model from the bovine esophagus, which is much more readily available in our country, was used as the first step of endoscopic submucosal dissection training. Twenty ESD procedures on the model were evaluated in terms of procedure time and muscle layer damage. Furthermore the first procedures performed by the endoscopist on humans were also evaluated, considering that they could provide insight into the effects of these procedures on humans.

The bovine esophagus, which was first 6 hours post-slaughter, was cleaned with tap water and 10% acetic acid (white vinegar) and placed in the prepared apparatus with a shoe box, 5-liter plastic bottle adjusted by cutting, 10 mililiter injection syringe, and gloves. An endoscopist who has not performed ESD before marked approximately 2-cm-diameter area with a Fujinon 530W gastroscope using a flexknife. The marked area was lifted with methylene blue and saline solution. Submucosal dissection was performed using the flexknife using the ERBE 300 cautery device, using the cut mode: effect 2 and watt 40, and the Swift coagulation mode: effect 4 and watt 40. In 20 procedures, the procedure time from the beginning of the marking to the completion of the dissection was assessed. After the procedure, the esophagus was dissected lengthwise to assess the presence of muscle damage at the procedure site. The presence of muscular damage in 20 ESD procedures performed on humans by the endoscopist were assessed retrospectively

In the bovine esophageal ESD animal model, the first dissection lasted 63 minutes, and two punctate superficial muscle damage was present. The mean time of the procedure was 30,15±14.8 min. After 10 procedures, a dissection of the same size took 23 minutes, and no muscle damage was detected. No muscular damage was detected in the 15 procedures following the 6th procedure. By the 20th procedure, dissection time had decreased to 14 minutes. In all procedures, the planned dissection area was extracted en-bloc. In the first 20 ESD procedures performed on humans by the endoscopist, who had used the bovine model 20 times, no complications were noted in 18 procedures. The mean diameter of the lesions was 25,1±12,6 mm and the time of the procedures (range 22-275 min) were variable according to the size, localization and fibrotic amount of tissue Muscular damage of 3 mm and 7 mm in diameter occurred in two patients. The damaged areas were closed with hemoclips.

In vitro and in vivo animal models constitute the first step in ESD training. The fact that the procedure time was reduced by two-thirds after 10 procedures and nearly 50 minutes after 20 procedures in the study suggests that it may be particularly useful for developing muscle memory with the use of the endoscope for mucosal incision and submucosal dissection. The in-vitro nature of the model and the absence of bleeding risk clearly played a role in this. While the presence of only one endoscopist in the study suggests that this model is not sufficient to determine its usefulness, similar findings have been obtained in in vitro porcine models.