Complications Following Circumferential Radiofrequency Energy Ablation of Barrett\'s Esophagus Containing Early Neoplasia

Abstracts

(n Z 2). Results: Initial attempts to perform precise sutured attachment with straight needles at flexible endoscopy were relatively unsuccessful. It was difficult to cannulate the attachment points and then penetrate the stomach at desired angle. The development of a novel flexible curved rotatable needle allowed rapid precise cannulation of the attachment points. With a laparoscopic view, the precise exit point of the needle could be identified before tissue puncture by seeing the transillumination from the gastroscope, tissue indentation from the needle and palpation using laparoscopic forceps. The distance from the aorta or other structures could be identified and adjusted during suturing. A stop on the thread behind the mucosal anchor allowed precise control of stitch length and avoided the need for thread tying. These devices and methods reduced the stitching time to place 8 full thickness attachment times to 20 minutes and increased precision and safety. In survival endoscopic and separate surgical attachment studies the 8 sutured attachment points were found to be holding the intragastric device in place at up to12 weeks without device detachments. Conclusion: A new flexible endoscopic suturing device and method was used with laparoscopic assistance for safe and secure anchoring of a food exclusion device to the cardio-esophageal junction.

S1474 Resection Area of 15 mm As Dividing Line for Choosing Strip Biopsy Or Endoscopic Submucosal Dissection for Mucosal Gastric Neoplasm Yasuharu Yamaguchi, Kenji Nakamura, Masao Toki, Kei Aoki, Shin-Ichi Takahashi Background: Endoscopic mucosal resection (EMR) is an effective alternative to surgical treatment for mucosal gastric neoplasm. Endoscopic submucosal dissection (ESD), a recently developed EMR method, has been reported to enable the en bloc resection for large lesions more frequently than conventional EMR methods such as strip biopsy (SB). There is no doubt that a tumor should be resected in en bloc fashion for accurate pathologic evaluation. But, it has been reported that ESD requires more time and is more often associated with a high incidence of complications than SB. Thus this study aimed to evaluate the size of resection area as dividing line for choosing SB or ESD for mucosal gastric neoplasm. Materials and Methods Study 1: Study 1 aimed to determine what size of resection area can be treated with SB as effectively as with ESD. We evaluated retrospectively 67 lesions (resection area less than 20 mm) treated with EMR (49 lesions; SB, 18 lesions; ESD) from July 01 to January 03. Study 2: From result of the study 1, using a 15 mm resection area as the dividing line between SB and ESD, study 2 will prospectively evaluate the relative effectiveness of SB and ESD from July 03. The subjects consisted of all patients undergoing EMR for mucosal gastric neoplasm. For the purpose of this study, SB was applied in lesions whose resection range was less than 15 mm including the marking (SB Group). In lesions which required 15mm or larger resection, ESD was performed for en bloc resection. We compared characteristics of lesions and outcomes of EMR between the two groups. Results Study 1:Of the patients with resection area less than 20mm, en bloc resection rate of SB and ESD for resection area less than 15mm were 90% (10/11) and 75% (6/8), respectively (p Z 0.22). The average resection time for such small lesions was significantly shorter with SB than ESD (12  5 vs. 70  30, p ! 0.01). For SB, the resection area less than 15mm satisfied both en bloc resection rate and resection time. Study 2: 90 lesions were prospectively assigned to SB group (36 lesions) and ESD group (54 lesions). The average neoplasm size was 9.0  3.9mm in the SB group and 19  11 mm in the ESD group. The average resection time was 12  6 min. in the SB group and 129  103 min. in the ESD group. The complete resection rate was 92% in the SB group and 83% in the ESD group (p Z 0.25). The complication rate was not significantly different between SB group and ESD group (11% vs 17%, p Z 0.12). During follow-up (median 46  10 months), one patient in each group had recurrent neoplasm. Conclusion: Gastric mucosal neoplasms which require only small (! 15mm) resection can be treated with SB, as effectively as with ESD.

S1475 Complications Following Circumferential Radiofrequency Energy Ablation of Barrett’s Esophagus Containing Early Neoplasia Roos E. Pouw, Joep J. Gondrie, Frederike G. Van Vilsteren, Carine Sondermeijer, Wilda Rosmolen, Wouter L. Curvers, Lorenza Alvarez Herrero, Fiebo J. Ten Kate, Kausilia K. Krishnadath, Paul Fockens, Bas L. Weusten, Jacques J. Bergman Background: Primary circumferential ablation (CA) and secondary focal ablation using radiofrequency energy has been proven effective in eradicating dysplasia and intestinal metaplasia (IM) in patients with Barrett’s esophagus (BE). The aim of this study was to evaluate complications following primary CA in patients with BE þ/ dysplasia, with and without prior endoscopic resection (ER) of visible abnormalities. Methods: All CA treatments were performed at our centre (July 2005present) under one or more EC approved protocols and prospectively entered into a dedicated database. According to each protocol, BE patients with high-grade dysplasia (HGD) or intramucosal cancer (IMC) first underwent ER to remove visible lesions and IMC, followed by primary CA 6 weeks later to treat residual BE þ/dysplasia. Results: 65 patients (50 men, median age 69 years, median Prague C4M6) were treated with primary CA; in 47 patients a prior ER had been performed. Four patients (6%) were hospitalized after CA for observation of fever (n Z 1), chest pain

www.giejournal.org

(n Z 2), and injury to a previous ER site followed by a negative contrast study (n Z 1). After conservative treatment and analgesics, all were discharged after 24-48 hours. In four patients (6%) a superficial laceration was observed immediately after CA. In one patient this was accompanied by a bleeding that was treated with adrenaline injection and hot biopsy forceps coagulation. All lacerations remained asymptomatic and did not require further treatment. Superficial lacerations all occurred in patients with prior ER (median of 2 resections, 33% of the circumference and 2.5 cm in length), and in whom the diameter of the ablation catheter exceeded the smallest diameter measured during preceding sizing of the esophageal inner diameter (ID). No lacerations were seen when an ablation catheter with a smaller diameter than the smallest measured ID was used (p Z 0.09). Five patients (8%) developed dysphagia, resolved with a median of 3 (IQR 15) endoscopic dilatations. These patients all had prior widespread ER (median of 3 resections, 50% of the circumference, and 2 cm length), multiple ER sessions, or a narrow esophagus at baseline. Conclusion: Primary CA using radiofrequency energy is safe in patients without prior ER. If ER is performed to resect visible abnormalities, complications at subsequent CA can be avoided by limiting the extent of the ER to 50% of the circumference and 2 cm in length, and by conservative selection of the diameter of the ablation catheter.

S1476 Novel Endoscopic Application of a New Flexible Fiber CO2 Laser for Esophageal Mucosal Ablation Sharmila Anandasabapathy, Dipen Maru, Sherry a. Klumpp, Rajesh Uthamanthil, Agatha Borne, Manoop S. Bhutani Background: The CO2 laser is one of the most widely used surgical tools in the aerodigestive tract because of its predictable penetration depth and minimal collateral damage due to efficient absorption of CO2 laser energy by tissue water content. Until recently, endoscopic use was limited by lack of an efficient delivery system. We evaluated the performance, efficacy, and safety of a novel, photonic band gap CO2 laser developed by OmniGuide Inc. for esophageal mucosal ablation in a swine model.Methods: Initial evaluation was performed on ex vivo swine esophageal tissue followed by endoscopic studies in freshly euthanized swine at 7, 10, 15, 20W power and 0, 1, 2, 5 & 10 mm distances, using continuous and pulsed current, to determine optimal performance settings. In an IACUC-approved protocol, 4 Yorkshire pigs were placed on omeprazole, fasted and underwent circumferential ablation of the distal 6 cm of the esophagus at 10W continuous current. The animals were survived 14 & 28 days to evaluate delayed tissue effects. Prior to necropsy, the proximal esophagus was ablated to evaluate the homogeneity of ablation and depth of injury immediately after single and repeat ablation with the laser. Results: Based on tissue & cadaveric studies, a setting of 10 W continuous current and a 5 mm working distance were deemed optimal. In the survival study, the treated pigs resumed a normal diet within 24 hours and showed no signs of dysphagia or weight loss. Pathology at 2 and 4-weeks revealed squamous reepithelialization with absent to minimal histologic evidence of injury and no strictures. In the freshly ablated areas, a single application of the laser produced complete transepithelial ablation of 96% of the treated surface area with a depth of injury extending to the muscularis mucosa in 68% and superficial (upper 10%) submucosa in 32%. With immediate sloughing and reapplication to the same area, 100% ablation was achieved with a similar depth of injury (34% superficial submucosa). One animal, ablated twice, had a focal injury to the superficial muscularis propria related to direct ‘wedging’ of the laser into the wall; however, no transmural injury was noted in any of the pigs. Conclusion: These porcine studies suggest a promising role for the flexible fiber CO2 laser in esophageal mucosal ablation. Homogeneous ablation was achieved with a predictable penetration depth (even with repeat application), and minimal tissue injury. These results warrant further evaluation of the laser as an ablative device for Barrett’s as it may overcome the limitations of currently available technologies including perforation, stricture, and lack of homogeneous ablation.

S1477 Investigation of the Thermal Tissue Effects of the APC Modes PULSED and PRECISE On Porcine Esophagus Ex Vivo and In Vivo Georg F. Kaehler, Ansgar Hieronymus, Rainer Grobholz, Mara N. Szyrach, Markus D. Enderle Background: Argon plasma coagulation (APC) is a monopolar, non-contact procedure, with the advantage of its relatively limited penetration depth. Reports on the penetration depth of APC are contradictory, and results vary between ex vivo and in vivo studies. The aim of the study was a systematic comparison of the tissue effects created by the 2nd generation APC modes PULSED Effect 2 and PRECISE in a porcine model ex vivo and in vivo. Material and MethodsAPC was carried out in vivo (variable probe to tissue distance due to peristalsis) during esophagoscopy in anesthetized pigs as well as ex vivo on explanted porcine esophagus (constant probe distance). Different power (PULSED APC Effect 2: 15 W, 25 W, 40 W) or effect settings (PRECISE APC: E2, E4, E6) and different application times (1, 2, 3 and 4 s) were used.ResultsIn vivo a well known, superficial tissue damage (type A), which extended maximally to the inner circular layer of the muscularis propria, and a new pattern of injury (type B), which extended to the outer longitudinal layer of the

Volume 67, No. 5 : 2008 GASTROINTESTINAL ENDOSCOPY AB145