AG

AA

bst

ract

s

Figure 1. Likelihood of PPI co-prescription according to number of risk factors for UGIC.

Su1780

CYP2C19*17 Gain of Function Mutation is Associated With the Developmentof Peptic Ulcer DiseaseCrispin O. Musumba, Diane Van Eker, Andrea L. Jorgensen, Eunice Zhang, NatalieO'Hara, Mark Pritchard, Munir Pirmohamed

INTRODUCTION: Studies show that single nucleotide polymorphisms (SNPs) in nonster-oidal anti-inflammatory drug (NSAID)-metabolising enzymes (mainly CYP2C9 & CYP2C8)may predispose NSAID-users to peptic ulcer disease (PUD) or upper gastrointestinal bleeding(UGIB). However, results to date have been inconclusive due to small sample sizes anddifferent study designs. METHODS: We hypothesised that the eight closely-linked clinicallyimportant SNPs in the CYP2C family of genes, namely CYP2C8*3 (rs11572080 &rs10509681), CYP2C8*4, CYP2C9*2, CYP2C9*3, CYP2C19*2, CYP2C19*3, andCYP2C19*17 may predispose to PUD via impaired NSAID metabolism as well as otherpotentially important mechanisms such as metabolism of arachidonic acid (AA) and proton-pump inhibitors (PPIs). Subjects endoscoped for suspected PUD/UGIB at 13 hospitals in theUK between 2005 and 2011 were recruited and interviewed using a structured questionnaire.Recruits were categorised as either NSAID-users (use within 2 weeks) or non-users (no usewithin 3 months). UGIB was defined as haematemesis, melaena or anaemia, and endoscopicstigmata of recent bleeding. H. pylori status was ascertained in most subjects. Followingextraction of genomic DNA, genotyping was performed by KBioscience Ltd (UK). Logisticregression analysis was used to test for association between each SNP and risk of PUD/UGIB. Interaction terms were introduced to determine whether any observed genetic effectswere influenced by factors including type of NSAID, PPI use and gender. RESULTS: Over-all,1246 white patients were recruited and categorised as follows: 485 (39%) PUD+/NSAID+;357 (29%) PUD+/NSAID-; 125 (10%) PUD-/NSAID+; 280 (22%) PUD-/NSAID-. Seven SNPswere included in the final analysis (CYP2C19*3 was monomorphic and excluded). All SNPswere in Hardy-Weinberg equilibrium. Logistic regression analysis of cases (PUD+; n=842)and controls (PUD-; n=405), assuming an additive mode of inheritance at each SNP, showedthat only CYP2C19*17 was significantly associated with PUD (p=0.006), with suggestionof an allele-dose effect, even on classifying cases as those using only CYP2C9/CYP2C8-substrate NSAIDs (p=0.005). Post-hoc analysis showed no interaction between CYP2C19*17and NSAID type, PPI use or gender. Subgroup analysis per ulcer type showed CYP2C19*17was significantly associated with risk of gastric ulcers (p=0.02), whilst only rs11572080 wasassociated with risk of duodenal ulcers (p=0.04). No SNPs were associated with UGIB.CONCLUSION: Possession of CYP2C19*17 allele is associated with PUD, especially gastriculcers, regardless of aetiology. We postulate that this could be through its effect on themetabolism of AA or other endogenous substances, leading to impairment of gastrointestinalmucosal defences. Further studies are needed to correlate the functional consequences ofCYP2C19*17 in the pathogenesis of PUD.

Su1781

Reactive Gastropathy and Correlation With Medications, Helicobacter pyloriand Gastric InflammationJennifer Hurrell, Byron L. Cryer, Rick Weideman, Ramiz Iqbal, Bert Little, Robert M.Genta

Introduction: Pathologists increasingly are using the term “reactive gastropathy” to describegastric biopsies showing histological abnormalities (especially foveolar hyperplasia) that areassumed to be due to topical injury from noxious chemical agents such as bile salts andNSAIDs. The pathogenesis of reactive gastropathy is not clear, however, and little is knownabout the contributions of H pylori and non-NSAID medications to the development of thislesion. This study was designed to explore these issues. Methods: At our VA Med Center,the study's two pathologists reviewed all gastric biopsy slides over 8 years from 2000-2007(5,664 cases), without knowledge of the original diagnosis, and classified them as: normal,chronic active gastritis with no H. pylori, H. pylori gastritis, chronic inactive gastritis, orreactive gastropathy. For each patient with reactive gastropathy, we identified two age-,gender- & biopsy date-matched controls with normal endoscopic gastric biopsies. Weexcluded patients with gastric polyps & gastric cancer. The electronic pharmacy records ofall study patients were reviewed to identify all medications (oral and parenteral) prescribed inthe one year preceding gastric biopsy. Relationships among reactive gastropathy, medicationexposures, H. pylori and other histologic categories were assessed by multivariate regressionanalyses. Results: We identified 489 patients with reactive gastropathy and 1147 controls.Mean age of our study population was 63 yrs (22-92 yrs). Histopathologic features of theoverall study group included H. pylori in 22%, CAG in 28%, intestinal metaplasia in 18%and atrophic gastritis in 3%. Gastric ulcers were found in 9% of subjects. The associationsof medications and histological features most and least strongly associated with reactivegastropathy are shown in the Table. We found no significant association between reactivegastropathy and the following medications: non-selective NSAIDs, glucocorticoids, PPIs andbisphosphonates. Also, age and gastric ulcer were not associated with reactive gastropathy.

S-502AGA Abstracts

Conclusions: Although reactive gastropathy has been assumed to result from topical chemicalinjury from potentially caustic agents like NSAIDs, we observed that the association ofreactive gastorpahty with NSAIDS is significantly associated with only COX-2 NSAIDS.Interestingly, we did find a significant association between reactive gastropathy and parenteralinsulin, a growth factor, which might contribute to the foveolar hyperplasia that characterizesthe lesion. Another hormonally-active therapy, spironolactone, which causes gynecosmastiaby blocking androgen receptors, also significantly affects reactive gastropathy. H. pyloriappears to protect against reactive gastropathy. Our findings suggest that reactive gastropathyis not solely a result of topical chemical injury.Risk of Reative Gastropahty

*parenterally-administered medication

Su1782

Effect of Proton Pump Inhibitors on the Discontinuation of Low-DoseAcetylsalicylic Acid in UK Primary CareElisa Martín-Merino, Saga Johansson, Luis A. Garcia Rodriguez

Introduction: Low-dose acetylsalicylic acid (ASA) is widely prescribed for the secondaryprevention of cardiovascular and cerebrovascular (CVD) events, but associated adversegastrointestinal events may lead to discontinuation. Proton pump inhibitors (PPIs) arerecommended for the prevention of ASA-associated adverse events in high-risk patients, yetlittle is known about the extent to which PPIs may impact on the rate of ASA discontinuationin clinical practice. Methods: Using the Health Improvement Network (a large UK primarycare database) a cohort of patients aged 50-84 years was identified who received at least 2prescriptions for low-dose ASA (75-300 mg/day) for secondary prevention of CVD between2000 and 2007 (N = 35 639). Discontinuation was defined as a period of ≥90 days aftercompletion of the last prescribed course of low-dose ASA during which no repeat prescriptionwas issued. A total of 11 729 patients discontinued low-dose ASA (cases) and 23 910continued therapy (controls) during a mean follow-up period of 2.5 years following firstlow-dose ASA prescription (start date). Current PPI use was defined as use on the indexdate (the date of discontinuation for cases and a random date for controls) or within theprevious 30 days, and PPI non-use was defined as no PPI use during the entire follow-upperiod. Data on peptic ulcer antecedents were ascertained prior to the start of follow-up.Multivariate adjusted logistic regression models were performed to evaluate the associationof low-dose ASA discontinuation with PPI use. The attributable risk was also calculated,with current PPI users who were also PPI users at start of follow-up (or in the following30 days) deemed protected, and non-users deemed not protected. Results: Current users ofa PPI were less likely to discontinue low-dose ASA therapy than non-users (OR [odds ratio]:0.92; 95% CI [confidence interval]: 0.87-0.97). The reduced likelihood of discontinuationassociated with current PPI use was seen among patients with peptic ulcer antecedents (OR:0.77; 95% CI: 0.64-0.92) and those without (OR: 0.92; 95% CI: 0.87-0.98). For every 100patients prescribed low-dose ASA, 27 discontinued ASA when co-prescribed a PPI, comparedwith 34 without concomitant PPI use: A 21.6% reduction in ASA discontinuation associatedwith PPI use. Conclusion: Our results suggest that co-prescription of a PPI in patientsreceiving low-dose ASA for secondary prevention of CVD will reduce the proportion ofpatients discontinuing therapy, thereby maximizing the cardioprotective benefit of low-dose ASA.

Su1783

Small Wheat Peptide Prevents Gluten-Dependent Maturation of DendriticCellsLuciana Giordani, Olimpia Vincentini, Maria Giovanna Quaranta, Cristina Felli, MarcoSilano, Marina Viora

Background: Celiac disease (CD) is a permanent immune-mediated enteropathy, triggeredin susceptible individuals, by gluten. Gluten is the alcohol soluble protein fraction of wheat,rye and barley. The main wheat gluten protein is gliadin. We have previously described theprotective activity of a peptide (QQPQDAVQPF, pDAV) towards celiac immune response.Dendritic cells (DC) are instrumental in the generation and regulation of immune responses.DC establish the balance between immunity and tolerance induction and are present indifferent compartments of the gut, including intestinal lamina propria, Peyer's patchesand mesenteric lymph nodes. Their priming ability is acquired upon maturation and ischaracterized by the activation of different transcriptional factors, leading to the modulationof genes involved in cytoskeleton rearrangement, antigen processing, control of migrationand regulation of inflammatory responses. Aim: The aim of the present study is to investigatethe effect of a peptic-tryptic (PT) digest of wheat gluten on normal human monocyte-derivedDC and the impact of pDAV in the modulation of PT-gluten-induced phenotypic andfunctional DC maturation. Methods: Immature DC were challenged In Vitro for 24 hourswith the peptic-tryptic (PT) digest of wheat gluten (500 μg/ml) and LPS as positive control.In some experiments immature DC were pre-treated with pDAV (100 μg/ml) or with a