Aims To identify the predictors of CD when presented with LD. Methods 215 LD patients had undergone prospective and systematic evaluation for CD and other recognised associations.The gold-standard diagnosis of CD was based upon the presence of HLA-DQ2 and/or DQ8, persistence or progression of LD following a gluten challenge, symptomatic improvement on a gluten-free diet, and no alternate cause found.Binary logistic regression models, adjusting for age and gender, were subsequently performed to compare presenting variables between CD and non-CD cases, and to determine their sensitivity, specificity, positive and negative predictive values (PPV and NPV).Results CD was diagnosed in 47 cases (22%) and non-CD in 168 cases (78%). There was no statistical difference in demographics, clinical symptoms (i.e. diarrhoea, weight loss, abdominal pain), anaemia or haematinics between the CD and non-CD group.Patients with CD, in comparison to non-CD, were significantly more likely to have a positive family history of CD (21.3% vs. 3.6%, OR 6.81; PPV 62.5%, NPV 81.4%, specificity 96.4%), positive HLA-DQ status (100% vs. 49.4%; PPV 36.2%, NPV 100%, specificity 50.6%), and presence of endomysial antibody [EMA] (49% vs. 0.6%, OR 159; PPV 96%, NPV 87%, specificity 99.4%); all p ≤ 0.001.A normal tissue transglutaminase antibody (TTG) level was seen in 29.8% CD and 82.7% non-CD cases (OR 0.086, p < 0.001; PPV 9.1%). There was no difference in the prevalence of TTG levels 1-2 x upper limit of normal (ULN) between the groups (29.8% CD vs. 14.3% non-CD; PPV 38%). However, TTG levels between 3-20 x ULN were significantly more prevalent in the CD group (31.9% vs. 3%; PPV 66.6% >87.5%),

IntroductionThe London Cancer Alliance is a group of 17 hospitals in North West and South London serving an estimated catchment area of 5.7 million. We have previously demonstrated that dedicated Barrett’s oesophagus (BE) surveillance lists improve dysplasia detection. LCA guidelines recommend dedicated surveillance lists, Prague reporting, dual pathology reporting of dysplasia and follow up intervals similar to national guidelines.MethodsOne month ‘snap shot’ audit of patients undergoing BE surveillance in June 2015. Structure of programme at each site was assessed with questionnaire. All patients undergoing endoscopy for BE were audited using endoscopy reporting software tools. New diagnoses or those referred with dysplasia for endotherapy were excluded. The Prague criteria, dysplasia detection rate and dual pathology reporting were assessed. Analysis was by independent t-tests for continuous variables and chi-squared tests for categorical variables.ResultsResponses were received from 13/17 sites. 6/13 ran dedicated surveillance lists and 5/6 had an active surveillance database. Dual pathology reporting for dysplasia was confirmed in 11/13 sites and BSG/LCA guidance adhered to in 12/13.A total of 137 patients underwent surveillance endoscopy. Results are shown in table 1. There was no difference in the mean number of procedures between sites with dedicated vs. non dedicated lists. Only 4/137 patients were diagnosed with dysplasia, with no significant difference between the 2 groups. Prague classification was significantly higher on dedicated lists (p < 0.0001). There was a significant difference between appropriateness of follow up between centres with active surveillance database vs. those without (p = 0.008).Abstract PWE-082 Table 1Study outcomesNon dedicated lists(n = 69)Dedicated lists (n = 68)P valueMean number of surveillance OGDs (±SEM)9.9 (2.1)11.3 (2.4)0.65Prague documentation68%97%0.0001Maximal length (±IQR)3.2 cm (0.29)3.9 cm (0.30)0.122≤ C0M256% (31/55)38% (26/68)0.17Dysplasia detection2.9% (2)2.9% (2)ns Database –ve Database +ve Appropriate follow up interval28/6268/750.0001ConclusionApproximately half the sites in the LCA have dedicated lists and/or active surveillance databases, with statistically significant higher Prague classification rates and appropriate choice of surveillance interval at follow up. These metrics could be adopted at a national level to assess quality of BE surveillance programmes. A larger national audit assessing utility of dedicated lists is warranted.Disclosure of InterestNone Declared