Tu1328 Endoscopic Submucosal Dissection for Barrett's Esophagus: Efficacy and Safety Results in a Large Tertiary Center Series of Patients

Abstract

0.001) were associated with non-response. Conclusion: RFA +/EMR +/APC is safe and effective for the treatment of dysplastic Barrett’s esophagus. Reflux esophagitis, stricture formation and greater length of Barrett’s esophagus were predictors of poor response. The majority of patients who either relapsed or progressed to cancer were, because of ongoing surveillance, able to be successfully treated with repeat RFA or surgery. Tu1328 Endoscopic Submucosal Dissection for Barrett’s Esophagus: Efficacy and Safety Results in a Large Tertiary Center Series of Patients Jean-Baptiste Chevaux*, Hubert H. Piessevaux, Anne Jouret-Mourin, Chun-Ping Ralph Yeung, Etienne Danse, Pierre H. Deprez Gastroenterology, Cliniques Universitaires Saint-Luc et Universite Catholique de Louvain, Brussels, Belgium; Pathology, Cliniques Universitaires Saint-Luc et Universite Catholique de Louvain, Brussels, Belgium; Radiology, Cliniques Universitaires Saint-Luc et Universite Catholique de Louvain, Brussels, Belgium Background & aims: Aims of this study were to assess efficacy, safety and long-term results of endoscopic submucosal dissection (ESD) in Barrett’s esophagus (BE). Materials and Methods: We analysed charts from 75 consecutive patients with BE treated by ESD from January 2007 through November 2013. ESD was performed in case of visible lesions, poorly lifting tumours, multifocal neoplasia, previous radiotherapy, and esophageal varices. Efficacy of ESD was assessed on the basis of rates of complete endoscopic and histologic remission concerning BE and safety on the rate of short and long-term complications. Results: Median age of patients was 68 years (IQR, 61-76) and 63 patients were male (85%). Median BE evolution was 41 months (IQR, 11-190) and median follow up was 18 months (IQR, 8-34). A mucosal defect was present in 92% of BE and classified 0-IIa (47%) and 0-IIc (42%) according to Paris classification. Median duration of procedure was 118 minutes (IQR, 100-146) and median circumference ablation was 75% (IQR, 66-80). Two procedures failed due to extensive submucosal fibrosis and a perforation. Five complications (6.6%) occurred during ESD including two haemorrhages and three perforations all treated endoscopically, no patients died. All patients were placed under proton-pump inhibitor regimen after ESD, 13 (17%) received oral corticosteroid therapy additionally. Median specimen size was 53 mm (IQR, 44-71) and good-to-moderate differentiation (G1/G2) was the more frequent encountered histologic type (75%). After ESD, endoscopic follow up was scheduled in all cases at 2 weeks. In case of increasing dysphagia or lumen narrowing, dilatation was performed on 45 patients (66%) with a median procedure number of 4 (IQR, 2-6). Additional endoscopic ablation of BE was compulsory for 40 patients (53%) and included mainly radiofrequency (28%). Median number of upper endoscopy per patient was 7 (IQR, 4-11) and median endoscopic procedure number for BE eradication was 2 (IQR, 1-2). Seven patients underwent surgery for vertical incomplete resection and/or undifferentiated lesions. At the latest control, BE endoscopic eradication was reached in 59%; histologic eradication of intestinal metaplasia in 58% and histologic eradication of neoplasia in 83%. Overall mortality was 8% but only one tumor-associated death was observed corresponding to an undifferentiated lesion with incomplete endoscopic resection in a ninety-year-old patient not suitable for surgery. Conclusion: we report the largest series of ESD performed in BE for specific indications such as mucosal or submucosal adenocarcinoma, multifocal neoplasia or poorly lifting epithelium. ESD appears to be safe and effective with complete remission of neoplasia similar to previously published data. Tu1329 Outcomes of Radiofrequency Ablation in Patients With Barrett’s Intramucosal Carcinoma Are Comparable to Those With Other Histologic Subsets of Be Christopher Blevins, Emmanuel C. Gorospe, Swapna Devanna, Cadman L. Leggett, Kelly T. Dunagan, Michele L. Johnson, Magdalen a. Clemens, Lori S. Lutzke, Kenneth K. Wang, Prasad G. Iyer* Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN Background and Aims: Esophageal endoscopic mucosal resection (EMR) followed by radiofrequency ablation (RFA) may be an effective treatment modality for Barrett’sassociated intramucosal cancer (IMC). Data on outcomes in this subset of BE related neoplasia particularly when compared to other subjects with BE undergoing endoscopic therapy are limited. We aimed to compare the outcomes of achieving complete remission of intestinal metaplasia (CRIM) and/or complete remission of dysplasia (CRD) in patients with IMC and other histologica subsets with BE undergoing RFA. Methods: We retrieved clinical, endoscopic and histopathologic data from a prospectively maintained database of patients undergoing treatment of Barrett’s esophagus. We included patients with baseline IMC as diagnosed by two expert gastrointestinal pathologists, in addition to those with HGD, LGD and NDBE. We compared the outcomes of those with IMC to those without (HGD, LGD and NDBE) undergoing RFA. IMC subjects were managed initially with staging using EUS, CT and PET scans, followed by EMR to achieve remission and then RFA to treated www.giejournal.org Vol residual BE. Successful endoscopic treatment was defined as achieving CRIM: absence of intestinal metaplasia in esophageal and GEJ biopsies from two consecutive endoscopic surveillance sessions. CRD was defined as the absence of dysplasia in two consecutive endoscopic surveillance sessions. Statistical analysis was conducted using standard statistical tests to compare means (SD) and proportions and time to event analysis. Results: We studied 339 patients who underwent RFA after EMR, with 55 (16%) having IMC and 284 (84%) without IMC : HGD (210, 61%), LGD (49, 14%) or NDBE (25, 7%) at the time of initial treatment. There was no difference in their baseline characteristics (age, sex, length of BE or length of hiatal hernia) as displayed in table 1. Among the IMC group, 34 (62%) of patients achieved CRD in a mean of 10.9 +/1.68 months and 43% achieved CRIM in a mean of 13.8 +/1.56 months following initiation of RFA. 212 (80%) of BE subjects without IMC reached CRD in an average of 13.1 +/1 months, and 139 (53%) achieved CRIM in an average of 18 +/1.3 months of RFA therapy. There was no significant difference between the proportion of patients who achieved CRIM (P Z 0.216) and the proportion of patients who achieved CRD (P Z 0.7927) in both groups. The time to reach CRIM (P Z 0.361; HR 1.45 [95% CI 0.865 to 2.303]) and time to reach CRD (P Z 0.3771; HR 1.29 [95% CI 0.721 to 2.176]) was also comparable between the two groups (Figure). Among subjects with IMC, 6 (32%) had recurrent IM with or without dysplasia in a mean time of 10.9 +/4.1 months, compared to 33 (24%) of subjects without IMC (pZ0.568). Conclusion: Our results suggest that RFA following EMR may have comparable clinical effectiveness for BE subjects with IMC compared to those without. Table 1. Comparison of baseline characteristics and outcomes of BE subjects with and without IMC treated with RFA