914 Patient-Controlled Analgesia With Inhaled Methoxyflurane Versus Conventional Sedation for Colonoscopy: A Randomized, Multi-Centre Trial Nam Q. Nguyen* 1 , Leanne Toscano 1 , James Moore 2 , Matthew Lawrence 2 , Richard H. Holloway 1 , William Tam 1 , Ian C. Roberts-Thomson 3 , Ilmars Lidums 3 , Venkataswamy N. Mahesh 1 , Carly M. Burgstad 1 , Tamara L. Debreceni 1 , Melissa Neo 1 , Mark Schoeman 1 1 Gastroenterology & Hepatology, Royal Adelaide Hospital, Adelaide, SA, Australia; 2 Colorectal Surgery, Royal Adelaide Hospital, Adelaide, SA, Australia; 3 Gastroenterology, The Queen Elizabeth Hospital, Adelaide, SA, Australia Discomfort during colonoscopy is usually managed with combined intravenous (IV) benzodiazepine(s) and opioid(s), which may lead to unnecessary sedation and associated adverse events. Patient-controlled analgesia with inhaled methoxyflurane (Penthrox ® ) has been used extensively in the community to manage the pain associated with trauma. Penthrox ® may be a more attractive alternative for managing the discomfort patients may experience during colonoscopy. Aims: To evaluate the efficacy, safety and post-operative management of Penthrox ® as an analgesic agent for performing colonoscopy as compared to conventional sedation with IV benzodiazepine and opioid. Methods: 103 patients who were referred for colonoscopy were randomized to receive either Penthrox ® (P: n48, 29M, 531.7yr) or IV midazolam and fentanyl (M&F: n55, 33M, 531.9yr) as a method of discomfort relief during colonoscopy. Patients with renal and liver diseases were excluded. Details on the degree of discomfort and anxiety before, during and after the colonoscopy were assessed using the visual analogue scale (VAS) pain score and STAI Y-1 anxiety score. Details on the performance of the colonoscopy as well as the occurrence(s) of adverse events were also documented. Vital signs and oxygen saturation during the procedure were monitored every 5 minutes. Data were analyzed using Fisher’s exact test and Student’s t-test. Results: Age and gender were similar between the two groups. In the sedation group, a mean dosage of 3.40.1mg midazolam and 83.62.9mcg fentanyl was used. There were no differences in: procedural success rate (sedation vs Penthrox ® : 54/55 vs. 47/48, respectively); hypotension during colonoscopy (4/55 vs. 7/48); tachycardia (3/55 vs. 4/48); caecal arrival time (9.60.8 vs. 8.40.7 min) and polyp detection/ polypectomy (19/55 vs. 18/48). Significant de-saturation (SaO 2 90%) was more common with sedation than with Penthrox ® (4/55 vs. 0/48; P0.05). Patients with Penthrox ® awoke earlier (1.50.4 vs. 11.12.5 min; P0.001) and were ready to be discharged earlier than those with sedation (36.02.2 vs. 56.62.5 min; P0.001). Neither VAS pain nor STAI Y-1 anxiety scores before, during or immediately after colonoscopy were different between groups. Conclusions: Patient-controlled analgesia with Penthrox ® for colonoscopy is as effective as conventional sedation for relieving pain and anxiety without influencing the procedural success and polyp detection rate. It appears safer with fewer events of oxygen de-saturation. Without the effects of sedation, patients with Penthrox ® analgesia are able to be discharged earlier which may facilitate early discharge and improve the work flow of endoscopy units. 915 Development of Novel Optical Technologies for Early Colon Carcinogenesis Detection via Nanoscale Mass Density Fluctuations: Potential Implications for Endoscopic Diagnostics Shailesh Bajaj* 1 , Ji Yi 2 , Andrew Radosevich 2 , Michael J. Goldberg 1 , Jeremy D. Rogers 2 , Laura K. Bianchi 1 , Eugene F. Yen 1 , Sudeep Upadhye 1 , Tat-Kin Tsang 1 , Beth Parker 1 , Hemant K. Roy 1 , Vadim Backman 2 1 GI/Medicine, NorthShore University HealthSystems, Evanston, IL; 2 Biomedical Engineering, Northwestern University, Evanston, IL Our group has been at the forefront of developing light scattering technologies (Nature 2000, Nat Med 2001, Gastro 2004, Clin Cancer Res 2006) and we have applied to cancer screening (reviewed in Gastro 2011). Our suite of technologies are sensitive to particle sizes from 50-500 nm thereby providing unprecedented insights into the nano-structural consequences of the genetic/epigenetic alterations in field carcinogenesis (Clin Cancer Res 2007, Gastro 2008, Cancer Res 2009). To maximize diagnostic performance, two critical issues need to be addressed, 1. Optimal depth to interrogate 2. Cells to target (crypt versus stromal). To address these issues we developed two new technologies. Depth was assessed via the tomographic functionality of low coherence enhanced backscattering spectroscopy (LEBS). To identify cells of origin, we invented inverse scattering optical coherence tomography (ISOCT) which allows derivation of the mass density correlation function . Methods: 82 patients undergoing screening /surveillance colonoscopy had two biopsies from the endoscopically normal rectal mucosa. Optical analysis was performed by investigators blinded to clinical data. The rectal biophotonic analysis was used to determine the difference in markers between controls (those with a negative colonoscopy) and patients harboring either advanced adenomas (1cm, 25% villous features or high grade dysplasia) or non-advanced adenomas (“effect size”(neoplasia- control)/SD). Results: Tomographic LEBS: We evaluated depth with a novel tomographic approach to LEBS. We evaluated penetration depth at progressively deeper LScs, and also measured LEBS enhancement which decreases early in field carcinogenesis (Cancer Res 2009). We noted the most robust effects appeared to be at superficial depths (Figure 1).ISOCT: In order to examine the epithelial versus stromal contributions, we used ISOCT to measure markers on crypt versus off crypt. We calculated all relevant tissue characteristics. Figure 1, we evaluated a derivative of spectral slope (a marker of submicron particle size heterogeneity) which we had previously shown to be related to neoplasia (Dis Col Rect 2009). We observed (figure 2) with the pseudo-color map that the greatest alterations were from the crypt, with the non- crypt site (stroma) showing more modest effects. Conclusions: We demonstrate that in early colon carcinogenesis, the nano-architectural changes are preferentially located in the cryptal structures at shallow (epithelial) depths. This study also underscores the power of combining two novel optical technologies, tomographic LEBS with ISOCT, for probing the nanoscale architectural alterations. In the future, these can be coupled with an endoscopically compatible fiberoptic probes in order to both detect dyplasia and also neoplastic risk through field carcinogenesis identification. 916 Accuracy of In Vivo Colorectal Polyp Discrimination Using Dual Focus High Definition Narrow Band Imaging Colonoscopy: A Randomized Controlled Trial; Preliminary Results Susan G. Coe*, Cristina Almansa, Julia Crook, Mihir K. Patel, Nancy Diehl, Estela G. Staggs, Vivian Ussui, Ernest P. Bouras, Andrew Keaveny, Michael F. Picco, Douglas Riegert-Johnson, Herbert C. Wolfsen, Michael B. Wallace Mayo Clinic, Jacksonville, FL Backgrounds: Advances in in-vivo imaging of colorectal polyps has opened the potential to replace histology for confirmation of small polyps with low malignant potential. Such a strategy could substantially reduce the cost of colorectal cancer prevention. The ASGE recently set threshold accuracy for polyp classification (90% NPV for adenoma), and surveillance interval predication ( 90% accuracy) before such a “diagnose and discard” strategy could be adopted.. Current methods such as chromoendoscopy and optical enhancement, including Narrow Band Imaging (NBI) have not generally met these thresholds. Zoom endoscopy has been shown to meet thresholds but is impractical in most settings. Aim: The aim of this study was to determine the accuracy of polyp classification comparing currently available HD-NBI systems (Olympus 180) to a pre-commercial next-generation system (Olympus 190) with NBI, pre-freeze image capture and dual focus high definition imaging (DFHD) in near mode or normal mode by a control button on the endoscope handle.Methods. Patients undergoing colonoscopy for screening or surveillance were randomized to standard HD-NBI imaging or DFHD-NBI colonoscopy. White light (WL) imaging was used in both arms for polyp detection. HD-WL and HD-NBI were used in the 180 arm, and near mode DFHD-WL and DFHD-NBI were used in the 190 arm for polyp classification using the NICE classification. The reference standard was the histological diagnosis by the clinical pathologist. The sample size was 600 patients randomized 1:1 to each arm. Six experienced staff endoscopists Figure 1. ISOCT Analysis of Rectal Spectral Markers Demonstrating Greatest Effect at the Crypt. Abstracts www.giejournal.org Volume 75, No. 4S : 2012 GASTROINTESTINAL ENDOSCOPY AB172