sequestrants were discontinued; loperamide was allowed as rescue therapy. Patients completed symptom diaries including stool frequency and Bristol Stool-form Scale (BSFS); a diarrhoea index ([stool frequency * mean BSFS] + loperamide use [weekly mg*3]) was calculated. Fasting serum FGF19 and total BA were measured before the first dose of OCA and after 2w treatment. Postprandial FGF19 and BA (6h area-under-curve, AUC) were determined after the first and last OCA dose. Data (expressed as

more »

... were analysed by Wilcoxon paired tests and Spearman correlation. Results OCA increased fasting FGF19 from 133 to 237 pg/ml (p = 0.007) at 2w. Most patients had an increase > 60% in fasting FGF19 and a large OCA first dose/postprandial response. Fasting BA reduced from 1.5 to 0.9 µmol/l (p = 0.13) and postprandial BA AUC was lower after the 2 w OCA treatment (from 4.9 to 3.0 µmol/l, p = 0.02). Clinical improvements were found in all patients, including in stool frequency (23 to 14/wk, p = 0.02), BSFS (5.15 to 4.34, p = 0.05) and the diarrhoea index (113 to 76, p = 0.005). The reduction in BA AUC (p = 0.02) and the increase in fasting FGF19 (p = 0.03) both correlated with the reduction in stool frequency. Symptoms of abdominal pain, urgency and bloating also tended to be less on OCA treatment. OCA was well tolerated and no adverse events were reported of clinical concern. Conclusion This study has shown for the first time that rational therapy with the FXR agonist OCA in PBAD is well tolerated and effective, stimulating serum FGF19 and reducing postprandial BA, resulting in clinical improvements in stool frequency and type. We propose larger, randomised, controlled trials of OCA. Introduction Lymphocytic duodenosis (LD) is defined by normal villous architecture and intraepithelial lymphocytes (IELs) > 25 per 100 enterocytes. Such patients should not be diagnosed with coeliac disease (CD), solely by histology, as recent studies have suggested other associations with LD. Coeliac serology (tissue transglutaminase [TTG] and/or endomysial antibodies [EMA]) may play a useful role although their diagnostic value in such settings is unknown. Aims To provide diagnostic outcomes in our expanding cohort of LD patients whilst also assessing the clinical utility of coeliac serology. Methods Two hundred patients with LD were investigated for CD and other known associations of LD, by means of revisiting the patient's history and recent investigations including the initial coeliac serology, followed by a combination of gluten challenge, HLA typing, repeat duodenal biopsies, and exclusion of infection/inflammatory bowel disease. In the absence of an alternative cause, a diagnosis of CD was based on the persistence or progression of LD on a gluten containing diet, the presence of HLA DQ2 or DQ8, and a clinical response to a gluten free diet. Results 150 female, 50 male, mean age 49, SD 16, age range 17-83 An identifiable association was found in 70% of patients -with CD (20%), NSAIDs (17%) and H,pylori (16%) accounting for the majority. In 30% no cause was found, although reassuringly 2/3rd normalised their histology. The role of coeliac serology in LD for diagnosing CD is shown in table 1. Conclusion As a single test, EMA has a greater diagnostic accuracy than TTG when assessing patients with LD. As a combination test, only the presence of both a positive EMA and a raised TTG has a 100% predictive value for CD.