Randomized Controlled Study of Detachable Loop Compared to Control, Cyanoacrylate, Or Sclerosant for Hemostasis of Bleeding Canine Gastric Varices

Abstract

Bleeding gastric varices are difficult to treat, particularly where cyanoacrylate (Glue) is not available for this indication. Glue for GV hemostasis in cirrhotics has also caused embolic complications that are often clinically significant. We are evaluating new potentially safer endoscopic treatments. Our specific aims were to compare efficacy, safety, & technical ease of endoscopic hemostasis for bleeding gastric varices (GV) by Olympus detachable loops (with & without sclerosant-ethanolamine + alcohol) vs. control, sclerosant (sclero), & glue. Methods: 10 adult dogs with prehepatic portal hypertension & fundal gastric varices (≥ moderate grade) were heparinized, entubated, & had EGD's. Five different GV/dog with positive Doppler ultrasound probe (DUP) signals were selected, bleeding was induced with a sclerotherapy needle, each site was randomized to control or 1 of 4 endoscopic treatments (glue, sclero, loop, or loop + sclero), & sites were mapped. Primary acute measures were total treatment time & ease of hemostasis (1 easy-4 difficult). Animals had daily PPI's (esomeprazole 40 mg) & weekly EGD's to grade GV size & blood flow (with DUP) + document 2° ulcer sizes, stigmata, & ulcer healing. Results: See Table for rates or means ± SEM, N = 10/group. No controls but all treatments had acute hemostasis. Ulceration rates were 0% for controls & 90% for all treatments. Glue plugged GV but did not scar or obliterate them or reduce GV size even when glue extruded. Other treatments caused significant scarring, obliteration &/or reduction of GV size. Conclusions: Loops with or without sclerosant compared to glue: 1. Were technically harder to use & had significantly longer treatment times. 2. Appeared to be safer in terms of 2° ulcer size, depth, major stigmata & ulcer healing rates. 3. Scarred & obliterated GV &/or significantly reduced GV size at 8 weeks. 4. Because they stopped GV blood flow, loops with sclerosant (or other agents) should not be associated with embolic complications. Supported in part by ASGE Research Award, Olympus Corp., Vascular Technology, & NIH K24 DK 02650. Tabled 1 Control Glue Sclero Loop Loop+Sclero Hemostasis time (sec) NA 70 ± 30 50 ± 9 284 ± 63 ∗ p < 0.05 vs. others 254 ± 29 ∗ p < 0.05 vs. others Ease of Hemostasis NA 1.5 ± 0.3 1.0 ∗ p < 0.05 vs. others 2.0 ± 0.3 1.7 ± 0.2 Mean ulcer diameter/depth (mm) NA 14.8 ± 2.4 ∗ p < 0.05 vs. others /2.8 ± 0.3 ∗ p < 0.05 vs. others 9.1 ± 0.8/2.3 ± 0.7 6.3 ± 0.9/1.2 ± 0.2 9.0 ± 1.2/1.4 ± 0.2 Major ulcer stigmata rate @ wks 1-8 (%) NA 78 ∗ p < 0.05 vs. others 44 ∗ p < 0.05 vs. others 11 11 Median wks to ulcer healing NA 10 ∗ p < 0.05 vs. others 4 1.5 2.5 GV grade ↓ @ 8 wks 0 0 −1.2 ± 0.2 ∗ p < 0.05 vs. others −1.8 ± 0.2 ∗ p < 0.05 vs. others −2.1 ± 0.1 ∗ p < 0.05 vs. others ∗ p < 0.05 vs. others Open table in a new tab