143 at medium volume (n = 4) & 118 at low volume centres (n = 15). There was no difference in entry criteria or demo- graphics between groups. CR-D & CR-IM at 12 months are no different between the groups (CR-D 86–90%, CR-IM 74– 81%). IM recurrence is significantly lower in high volume centres (16.1% vs 20.3% and 19.2%, Log Rank p < 0.001) but dysplasia recurrence is no different (Log Rank p = 0.12). Rescue EMR was performed less frequently in medium vol- ume centres (0% vs high 5.3% and low volume 10%, p = 001). Conclusion Endotherapy for Barrett’s dysplasia is highly effec- tive whatever the centre volume. The rescue EMR rate in medium volume centres is unexplained. Despite lower IM recurrence in high volume centres, dysplasia recurrence rates are not significantly different. Caseload volume of a centre in the UK Registry does not appear to affect outcome. Disclosure of Interest None Declared PWE-077 RESIDUAL INTESTINAL METAPLASIA AFTER SUCCESSFUL ENDOSCOPIC THERAPY FOR BARRETT’S RELATED NEOPLASIA CONFERS HIGHER LONG TERM RISK FOR DISEASE RECURRENCE, ON BEHALF OF THE UK RFA REGISTRY 1,2 G Lipman * , 2 A Gupta, 3 JM Dunn, 2 D Morris, 4 H Smart, 5 P Bhandari, 6 RP Willert, 7 G Fullarton, 7 AJ Morris, 8 M Di Pietro, 9 C Gordon, 10 I Penman, 11 H Barr, 12 P Patel, 12 P Boger, 13 N Kapoor, 14 BS Mahon, 15 J Hoare, 16 R Narayanasamy, 16 DO’Toole, 17 Y Ang, 18 A Veitch, 19 D Nylander, 20 A Dhar, 21 K Ragunath, 22 A Leahy, 22 M Fullard, 1,2 R Haidry, 1,2 LB Lovat, on behalf of The UK RFA Registry. 1 Division of Surgery & Interventional Science, University College London; 2 University College Hospital NHS Foundation Trust; 3 Guy’s and St Thomas’ NHS Foundation Trust, London; 4 Royal Liverpool University Hospital, Liverpool; 5 Queen Alexandra Hospital, Portsmouth; 6 Central Manchester University Hospitals NHS Foundation Trust, Manchester; 7 Glasgow Royal Infirmary, Glasgow; 8 Addenbrookes Hospital, Cambridge; 9 Royal Bournemouth Hospital, Bournemouth; 10 Royal Infirmary Edinburgh, Edinburgh; 11 Gloucestershire Hospital NHS Trust, Gloucestershire; 12 Southampton University Hospital, Southampton; 13 Aintree University Hospital, Liverpool; 14 Queen Elizabeth Hospital, Birmingham; 15 St Mary’s Hospital, London, UK; 16 St James Hospital, Dublin, Ireland; 17 Salford Royal Foundation NHS Trust, Manchester; 18 Wolverhampton NHS Trust, Wolverhampton; 19 Newcastle upon Tyne Hospitals NHS Foundation Trust, Newcastle; 20 County Durham Hospital, County Durham; 21 Nottingham University Hospital NHS Trust, Nottingham; 22 West Hertfordshire Hospitals NHS Trust, Watford, UK 10.1136/gutjnl-2016-312388.322 Introduction Endoscopic resection (ER) followed by Radiofre- quency ablation (RFA) is the first line treatment for neoplastic Barrett’s oesophagus (BE). Metachronous neoplasia after focal eradication of disease is ~20%. We examine data from the UK registry of 28 centres to establish if residual metaplastic BE carries a risk of disease recurrence. Methods Visible lesions were removed by EMR. Patients then underwent RFA 3 monthly. Biopsies were taken at 12 months to assess treatment success with repeat biopsies every 6–12 months thereafter.. Dysplasia recurrence was compared in patients who had complete reversal of BE and neoplasia (CR- IM) to those in whom dysplasia alone was eradicated (CR-D only). Residual BE was confirmed with visible columnar epi- thelium proximal to the OGJ with biopsies showing IM. Results 517 patients achieved CR-IM & 96 patients achieved CR-D only after 12 months treatment . Sex & ER rates were not significantly different between groups. The CR-D only group were older (mean age 70 vs 67, p < 0.01) and had longer initial BE (mean length 6.2 cm vs 4.7 cm, p < 0.0001). Mean residual BE length was 1.4 cm. At median follow up 32 months, more patients were disease free in the CR-IM group (96% vs 89%) and Kaplan Meier statistics demonstrated an improved predicted 6 year neoplasia free survival in the CR-IM group (90% vs 84% log rank p 0.0015). Most recurrences occurred within 3 years of follow up. Conclusion Endotherapy should aim to clear neoplasia and underlying metaplastic BE to improve long term outcome. Patients with CR-D but not CR-IM at the end of treatment have an increased risk of neoplasia recurrence. This may have implications for post treatment surveillance intervals. Disclosure of Interest None Declared PWE-078 MAGNIFICATION ENDOSCOPY WITH I-SCAN IMAGING AND ACETIC ACID CHROMOENDOSCOPY IN BARRETT’S OESOPHAGUS IMPROVES NEOPLASIA DETECTION 1,2 G Lipman * , 3 R Bisschops, 4 J Ortiz-Fernández-Sordo, 1 R Sweis, 5 JM Esteban, 1,2 LB Lovat, 4 K Ragunath, 1,2 R Haidry. 1 Department of Gastroenterology, University College Hospital NHS Foundation Trust; 2 Division of Surgery & Interventional Science, University College London, London, UK; 3 University Hospitals Leuven, KU Leuven, Belgium; 4 Nottingham Digestive Diseases Biomedical Research Unit, Nottingham, UK; 5 Hospital Clínico San Carlos, Madrid, Spain 10.1136/gutjnl-2016-312388.323 Introduction Barrett’s oesophagus (BE) is the pre-cursor for oesophageal adenocarcinoma. Endoscopic surveillance is per- formed in BE patients to detect dysplasia as an early treat- ment target. Current surveillance relies on random quadrantic biopsies every 1–2 cm through the BE with targeted biopsies for areas of suspicion. This strategy samples less than 5% of the BE mucosa. We present a novel endoscopic classification system utilising magnification chromo-endoscopy with i-Scan (PENTAX HOYA, Japan) image enhancement technology and acetic acid to improve dysplasia recognition in BE. Methods High definition (HD) video recordings of suspicious lesions were collected from patients with non-dysplastic (ND- BE) and dysplastic (D-BE) BE undergoing endoscopy at a high volume tertiary centre. Lesions were recorded with magnifica- tion endoscopy in all i-Scan modes both before and after application of 2% acetic acid (ACA) before sampling with biopsy forceps or endoscopic mucosal resection to confirm the histological diagnosis. Six expert endoscopists scored videos using a previously validated mucosal (M) and vascular (V) classification system. Normal mucosa was defined as regular circular or villous pits (M1) and abnormal mucosa defined as distorted or irregular pits or featureless mucosa (M2). Normal vascular pattern was defined as regular and uniform vessels (V1) and abnormal vascular pattern was defined as irregular, dilated corkscrew vessels (V2). Dysplasia was classified if the lesion was felt to be either M2 or V2. Results 63 lesions (36 D-BE, 27 ND-BE) were recorded (30 before ACA and 33 after ACA) for analysis. Experts’ average accuracy for dysplasia prediction was 67.8% pre ACA and 75.9% after ACA (p = 0.01). ACA improved the sensitivity of our novel classification system for neoplasia detection from 81% to 88% (p = 0.04). Interobserver Kappa values were 0.253 pre ACA and 0.369 after ACA. Conclusion Experts can diagnose D-BE in up to three-quarters of cases using i-Scan magnification endoscopy with acetic acid. Abstracts A176 Gut 2016;65(1):A1–A310