Endoscopic submucosal dissection for large early squamous cell carcinoma—traction assisted methods

Abstract

: Endoscopic resection is a minimally-invasive treatment for superficial esophageal tumors compared with surgery or chemoradiation therapy. Endoscopic submucosal dissection (ESD) is a well-established method with accurate histological evaluation and favorable procedural outcomes. However, ESD requires a high level of skill, and is therefore technically challenging and time-consuming. Traction-assisted ESD has been introduced to facilitate ESD and provides adequate submucosal visualization and satisfactory tissue traction. We reviewed the scientific literature in English to evaluate the efficacy of traction-assisted ESD for esophageal lesions, including the clip-with-thread (CT) method and the submucosal tunneling (ST) method. The CT method is a simple and affordable technique that uses commercially available hemoclips. Two randomized controlled trials and two retrospective studies showed that the CT method resulted in shorter procedure times and reduced local injection compared with conventional ESD, and no cases of perforation. The ST method does not require specific devices and facilitates the procedure by securing a stable submucosal visual field and maintaining a submucosal liquid cushion. Two retrospective studies reported shorter procedure times with similar en bloc resection rates and complete resection rates using the ST method compared with conventional ESD. Although a combination of the ST method with the CT method seems to be effective for large lesions, its efficacy and safety should be confirmed by a large-scale study. In the future, robotic traction has a great potential to be a breakthrough for esophageal ESD, providing appropriate and multi-directional traction to the lesions via an operator-controlled robotic arm. There is no doubt that traction assistance is key to facilitating esophageal ESD. Further studies are needed to elucidate the best method from the perspective of efficacy, safety, and cost.