Sa1929 VERIFICATION OF USEFULNESS OF 3-DIMENSIONAL ENDOSCOPY IN ESOPHAGEAL ESD OF AN EX VIVO ANIMAL MODEL

Abstract

Because 2-dimensional (2D) endoscopy cannot provide depth information, endoscopists feel difficulties during ESD procedure when an endoscope perpendicularly faces to the muscular layer. 3-dimensional (3D) visualization offers better depth recognition. Therefore, 3D rigid endoscopy has already been clinically introduced in surgical fields to enable more accurate and safer procedures. To explore a feasibility of 3D flexible endoscopy, we conducted a study to compare 2D and 3D endoscopy in the performance of esophageal ESD using a newly developed 3D flexible endoscope (Olympus scientifics, Tokyo, Japan). 2D observation is easily changed to 3D observation with a switching button. Isolated porcine esophagus was used for ESD. A virtual target lesion was made by marking a 15mm-diameter area with a dual-J knife (Olympus scientifics). Hyaluronic acid solution was used for injection. Six endoscopies (3 experts and 3 trainees) participated in this study. Endpoints evaluated were 1) en bloc resection rate, 2) procedure time of submucosal injection and incision/dissection (sec), incision/dissection speed [resected area (mm2) / procedure time (sec)], 3) numbers of technical complications (perforation, muscle layer damage and sample damage), and 4) degree of sense of security during ESD procedure, degree of fatigue and eye-strain assessed by a visual analog scale (VAS). 1) En bloc resection rates were 100% in both 2D and 3D endoscopy. 2) Procedure time of submucosal injection and incision/dissection were equivalent between 2D and 3D endoscopy in both experts and trainees. 3) Incision/dissection speeds (mm2/sec) were equivalent between 2D and 3D; those of experts were 0.38±0.06(mm2/sec) in 2D and 0.39±0.036(mm2/sec) in 3D, and those of trainees were 0.22±0.003(mm2/sec) in 2D and 0.22±0.02(mm2/sec) in 3D. 4) Numbers of technical complications in 2D (3.5±4.09, mean±SD) tended to be higher than those in 3D in trainees (1.33±2.80) (P=0.063). Those in experts were equivalent between 2D and 3D. 5) Degrees of sense of security in 3D (3.67±0.82, mean±SD) were significantly higher than those of 2D (2.67±0.52) in trainees (p=0.041), but those were equivalent between 2D and 3D in experts. Degrees of eye-strain in 3D (3.00±0.00) were significantly higher than those in 2D (2.17±0.41) in trainees, but were equivalent in experts. Degrees of fatigue showed no significant difference between 2D and 3D in both trainees and experts. In esophageal ESD of an ex vivo animal model, 3D endoscopy reduced technical error rate and improved sense of security during procedure in trainees, but not in experts. 3D flexible endoscopy might be feasible and improve qualities of ESD. Eye-stain could be a limitation of 3D endoscopy, and further studies and developments were required to elucidate possibilities of 3D technology in flexible GI endoscopy.