The Efficacy of an Endoscopic Grasp-and-Traction Device for Gastric Endoscopic Submucosal Dissection: An Ex Vivo Comparative Study (with Video).
- Author:
Dirk W SCHOLVINCK
1
;
Osamu GOTO
;
Jacques J G H M BERGMAN
;
Naohisa YAHAGI
;
Bas L A M WEUSTEN
Author Information
- Publication Type:Comparative Study ; Original Article
- Keywords: Endoscopic submucosal dissection; EndoLifter; Stomach neoplasms; Gastric mucosa; Swine
- MeSH: Gastric Mucosa; Stomach; Stomach Neoplasms; Swine
- From:Clinical Endoscopy 2015;48(3):221-227
- CountryRepublic of Korea
- Language:English
- Abstract: BACKGROUND/AIMS: To investigate whether the EndoLifter (Olympus), a counter-traction device facilitating submucosal dissection, can accelerate endoscopic submucosal dissection (ESD). METHODS: Two endoscopists (novice/expert in ESD) performed 64 ESDs (artificial 3-cm lesions) in 16 ex vivo pig stomachs: per stomach, two at the posterior wall (forward approach) and two at the lesser curvature (retroflex approach). Per approach, one lesion was dissected with (EL+) and one without (EL-) the EndoLifter. The submucosal dissection time (SDT), corrected for specimen size, and the influence of ESD experience on EndoLifter usefulness were assessed. RESULTS: En bloc resection rate was 98.4%. In the forward approach, the median SDT was shorter with the EndoLifter (0.56 min/cm2 vs. 0.91 min/cm2), although not significantly (p=0.09). The ESD-experienced endoscopist benefitted more from the EndoLifter (0.45 [EL+] min/cm2 vs. 0.68 [EL-] min/cm2, p=0.07) than the ESD-inexperienced endoscopist (0.77 [EL+] min/cm2 vs. 1.01 [EL-] min/cm2, p=0.48). In the retroflex approach, the median SDTs were 1.06 (EL+) and 0.48 (EL-) min/cm2 (p=0.16). The EndoLifter did not shorten the SDT for the ESD-experienced endoscopist (0.68 [EL+] min/cm2 vs. 0.68 [EL-] min/cm2, p=0.78), whereas the ESD-inexperienced endoscopist seemed hindered (1.65 [EL+] min/cm2 vs. 0.38 [EL-] min/cm2, p=0.03). CONCLUSIONS: In gastric ESD, the EndoLifter, in trend, shortens SDTs in the forward, but not in the retroflex approach. Given the low numbers in this study, a type II error cannot be excluded.
