500 The “Speedboat”: a New Multi-Modality Instrument for Endoscopic Resection in the Gastrointestinal Tract Brian P. Saunders* 1 , Zacharias P. Tsiamoulos 1 , Leonidas a. Bourikas 2 , Paul D. Sibbons 3 , Christopher P. Hancock 4 1 Wolfson unit for Endoscopy, St Mark’s Hospital and Academic Institute, London, United Kingdom; 2 Department of Gastroenterology, University Hospital of Crete, Heraklion, Greece; 3 Department of Surgical Sciences, Nortwick Park Institute for Medical Research, London, United Kingdom; 4 Creo Medical Ltd and School of Electronic Engineering, Bangor University, Chempstow and Bangor, United Kingdom Background: Large sessile or flat intestinal lesions (2cm) are optimally removed en-bloc for more accurate histology and for completeness of resection. Current submucosal dissection devices are technically challenging to use, resulting in long and sometimes incomplete procedures with a high risk of major complication. We describe, for the first time, a simple to use, multi-modality endoscopic device (‘Speedboat’, Creo Medical Ltd) for wide-field, en-bloc mucosal resection. Aims and Methods: The ‘Speedboat’ (Creo Medical Ltd) cuts in forward, lateral and oblique planes using bipolar radio frequency (RF) cutting, provides haemostasis with microwave (MW) coagulation and incorporates a retractable needle for submucosal injection and tissue irrigation. The instrument blade has an insulated ‘hull’ to prevent thermal injury and the device catheter is partially torque stable allowing rotation and orientation of the hull to protect the underlying muscularis mucosa. The electrosurgical generator used to deliver the cutting and haemostasis power to the blade is comprised of 2 power sources, one operating at 400KHz (RF) and the other at 5.8GHz (MW). Speedboat submucosal dissection (SSD) technique was performed and video recorded on 4 consecutive 60kg pigs. Mucosal areas to be resected were marked prior to submucosal injection. The mucosa was then circumferentially incised and resected by SSD. The time taken to complete resection, complications encountered and power settings used were recorded. Immediately after the procedure, the animals were euthanized, and the resection defects measured and assessed histologically. Results: Eight consecutive resections were performed (2 per animal), 7 in the colorectum and 1 in the antrum of the stomach. The median time to complete a resection was 37 minutes range (30-60 minutes) using RF cutting with 24W, voltage circa 300Vrms. Median defect size (longest diameter) was 53.5mm, range 40-80mm. Microwave coagulation was applied for either minor bleeding or visible vessels on 32 occasions (average power 7.5W). An endoclip was used once to control arteriolar bleeding but no other haemostatic device was required. There were no perforations and histology (picrosirius red and hematoxylin & eosin staining) showed an intact and viable muscle layer with some remaining submucosa in all cases. Conclusions: This initial evaluation of the Speedboat suggests that it may facilitate rapid and safe en-bloc mucosal resection in the colorectum and gastric antrum. 501 A Novel Viscous Dissecting Gel Is Safe, Simple and Rapid for Endoscopic Submucosal Dissection: a Prospective, Randomised Controlled Trial Payal Saxena*, Yamile Haito Chavez, Ali Kord Valeshabad, Venkata S. Akshintala, Brian W. Simons, Vikesh K. Singh, Minmin Zhang, Faming Zhang, Eun Ji Shin, Kathleen Gabrielson, Anne Marie Lennon, Jie Peng, Patrick I. Okolo, Marcia I. Canto, Anthony N. Kalloo, Mouen Khashab Johns Hopkins Medical Institute, Baltimore, MD Background: Endoscopic submucosal dissection (ESD) can resect en bloc large early neoplastic lesions but it involves specialized technical skills and training, increased procedure time and higher complication rates compared to endoscopic mucosal resection (EMR). We have shown that a new viscous gel (VG) (Cook Medical) facilitated ESD due to its submucosal (SM) dissecting properties. Aim: To compare procedure time of ESD using standard techniques (ESD-S) with ESD using VG (ESD-G) of simulated 3cm gastric lesions in a porcine model. Secondary goals were to assess en bloc resection rates, short and long term complications, and histological injury score on day 28. Methods: 3cm gastric simulated SM lesions were created by SM injection of saline with 0.3% indigo carmine, followed by placement of 10mm embolization coils into the bleb using 19-gauge needles. Randomization to ESD-G or ESD-S was performed by sealed envelope allocation. All procedures were performed by a single endoscopist with 12 month experience in ESD. ESD-G was performed by injecting the gel into the SM bleb using a 19-G needle and a custom-made injector apparatus. A circumferential SM incision was made around the periphery of the lesion with a triangle tip (TT) knife (Olympus) using pure cut mode (Erbe), followed by snare (33mm diameter) resection of the lesion. SM dissection was not needed due to “auto-dissecting” properties of the gel (Figure 1). ESD-S was performed using standard techniques with a TT knife. Procedural and resection times were recorded. All pigs were survived until necropsy at day 28. Pathologists were blinded to ESD technique. Results: 12 simulated gastric lesions were equally randomized to ESD-G and ESD-S. Mean lesion size was similar in both groups (40.7mm vs. 37.7mm, p=0.69). Average of 17mL of VG was injected during ESD- G. En bloc resection rate was 100% in the ESD-G group as compared to 67% in ESD-S group (p=0.45). There was one episode of intra-procedural bleeding requiring endoscopic hemostasis in the ESD-S group, but none in the ESD-G group. Mean total procedural time (13.5 min vs. 28.7 min, p=0.005) and mean resection time (5.5 min vs. 23.8 min, p=0.001) were significantly shorter in the ESD-G group. VG resulted in 53% and 77% reduction in total procedure time and resection time, respectively. There were no delayed complications in either group. Histology confirmed complete mucosal resection without perforation or penetration of the muscularis propria in all cases. The mean overall histologic injury score was similar between both groups (Table 1). Conclusion: ESD-VG of large gastric lesions can dramatically reduce procedural time and resection time. VG is safe and easy to use. It eliminates the need for time-consuming SM dissection and simplifies ESD. Table 1. Histologic injury comparison between ESD-G and ESD-S groups ESD-G (n6) ESD-S (n6) p value Total injury score (mean) 4 5 0.15 Edema score (mean) 1=minimal edema in submucosa 2=moderate edema in submucosa 3=moderate to severe edema in multiple 1.17 1.5 0.26 Hyperemia/vasodilation score (mean) 1=minimal 2=moderate 3=severe 1.5 1.7 0.60 Hemorrhage score (mean) 1=minimal or focal hemorrhage 2=moderate or multifocal hemorrhage 3=severe or widespread hemorrhage 1.3 1.8 0.09 Fig 2. Images of endoscopic hand-sewn suturing. After derivering the needle and string through a overtube, a mucosal defect was firmly sutured with a prototype of a needle holder. Abstracts AB155 GASTROINTESTINAL ENDOSCOPY Volume 77, No. 5S : 2013 www.giejournal.org