Relative performance of four second generation flexible endoscopic notes closure techniques (FECT's)

Aim: To assess efficacy, difficulty and reproducibility of four FECTs, qualified by a) performability with available devices through a flexible endoscope; b) feasibility and potential already shown; and c) lack of prior formal comparisons.

Methods & Procedures: Two-phase study. Phase I: 14 endoscopists closed NOTES access perforations (needle-knife& 15-mm-balloon dilation) on 52 pig stomachs/colons (half each) isolated and frozen within 2 weeks. All 14 operators carried out one closure each using T-Anchors (Cook) or TA, Clips (Boston Scientific) on Loops (Olympus) or COL, and Overclips (OTSC, circular clips mounted rubber-band-like on a cap at the endoscope tip, from Ovesco). 10 carried out Submucosal Endoscopy with Mucosal Flap (SEMF) closures, created by saline injection, needle-knife entry, and balloon dissection. Standard (OTSC & SEMF) or double channel (TA & COL) gastroscopes were used. Air leakage was tested and data collected for each closure: success/failure (1^st& 2^nd attempt); procedural time (min); leak pressure (mm Hg); and blinded complexity assessment (3-point scale). Comparisons with univariate analysis, Scheffé test, and Chi squared where appropriate, to identify the best performing FECT. With this, 10 additional closures were carried out acutely in Phase II on the stomachs and colons of 5 large white pigs (20 kg). After autopsy, leak testing was performed, and both leak pressure and complexity score were compared with Student's test to those in Phase I.

Results: The mean ± SEM leak pressures and procedure times for TA/COL/OTSC/SEMF were 26.9 ± 4.2/13.5 ± 2.0/42.4 ± 7.2/19.5 ± 7.2 (p = 0.02, Scheffé OTSC Vs COL only significant), and 33.6 ± 2.9/ 21.0 ± 1.4/18.6 ± 1.6/59.8 ± 8.3 (p = 0.00, Scheffé OTSC Vs COL only non-significant). Whereas all but SEMF succeeded at the second attempt, initial success for TA/COL/OTSC/SEMF was 78%/93%/93/ 50% (p = 0.032) and the corresponding (easy-average-difficult) score for those FECTs was (4-7-3)/(3-4-7)/(7-7-0)/ (1-0-9) (p = 0.001). Thus, OTSC with the highest leak pressure, lowest complexity score, highest initial success rate and lowest procedure time (these last two on par with COL) was identified as the best performing FECT. In-vivo OTSC leak pressures was 79.5 ± 10.4, higher than ex-vivo (p = 0.06), with a no different complexity score.

Conclusions: Overclips outperformed previously reported FECTs in sequential porcine ex-vivo and in-vivo closures carried out by 14 different operators.