AGA Abstracts present study, we aimed to characterize clinical features of patients diagnosed at our institu- tion with lymphocytic gastritis. Methods: Using a search of our electronic pathology record for the term "lymphocytic gastritis" spanning the years 1997-2014, we identified cases with this term and reviewed clinical features including indication for biopsy, co-morbid diseases, and medication use. LG was defined as 25 intraepithelial lymphocytes/100 epithelial cells, regardless of inflammation in the lamina propria. Results: 72 patients were confirmed to have lymphocytic gastritis by pathology report. The median age was 45 (range from 3 to 90) and 61% were female. The most common indication for biopsy was iron deficiency anemia/suspected celiac disease. Of the 72 patients, 45 (63%) had celiac disease, 8 (11%) had H. pylori , 2 (2.8%) had celiac disease and H. pylori , 1 (1.4%) had HIV, 4 (5.6%) had lymphoma, and 19 (26%) had no known associated condition. Current medication use was available for 45 patients. The most common medications were: ACE/ARB (14%), including olmesartan (3 patients, 4.2%), PPIs (11%), and SSRIs (5.6%). In the patients without known association with LG the most common medications were: PPIs (21%), ACE/ARBs (16%), and SSRIs (5.3%). 1 patient (5.2%) was taking an ACE and a PPI. 11 patients had colonic biopsies with 4 patients (36%) having evidence of microscopic colitis. Conclusion: In this series of 72 patients with LG, celiac disease was found to be the most common underlying disease process. LG was also seen in patients on PPIs, SSRIs, and ACE/ARBs (including olmesartan). In patients without a known cause of LG, PPIs were the most common medica- tion used. Concomitant microscopic colitis was seen in a proportion of patients. Sa1292 Concomitant Celiac Disease Does Not Modify the Response to Treatment With Bismuth Subsalicylate in Patients With Microscopic Colitis Clayton M. Spiceland, Nicole M. Gentile, Sahil Khanna, Darrell S. Pardi BACKGROUND: Patients with microscopic colitis (MC) have an increased risk of celiac disease (CD) compared to the general population. Bismuth subsalicylate (BSS) is used as a treatment for diarrhea in patients with MC. We evaluated a cohort of patients with CD and MC who received BSS treatment and compared their response to a cohort of patients with MC without CD. METHODS: Adult patients with MC treated with BSS at our institution were identified over a ten-year period from 12/31/2001 to 12/31/2011. The electronic medical record was searched to identify patients with either lymphocytic colitis (LC) or collagenous colitis (CC). Active disease was defined as the presence of diarrhea along with compatible biopsies taken from normal appearing colonic mucosa. Pathology reports and tissue transglutaminase serologies were reviewed to identify concomitant CD diagnoses. BSS dose regimens varied from 6 to 9 tablets per day. Response to BSS treatment was defined as complete response (resolution of diarrhea at 8 weeks), incomplete response (>=50% improvement from baseline), or no response (<50% improvement). Response to BSS in patients with MC and CD was compared to an age- and gender-matched cohort of patients with MC without CD. RESULTS: 64 patients with MC treated with BSS were identified; 10 patients (15.6%) had concomitant CD. The patients with concomitant CD had a median age of 48.5 years (range 26-85) and 100% were female; 60% had CC and 40% had LC. 90% were diagnosed with CD prior to the diagnosis of MC. All patients had biopsy proven celiac disease, 70% had concomitant positive serology, and 100% responded to a gluten free diet at time of CD diagnosis. Overall, complete response to BSS was seen in 40%, partial response in 30%, and no response in 30%. The response to BSS in the matched MC cohort without CD was complete in 60%, partial in 20%, and none in 20% (p=NS). CONCLUSIONS: In this cohort of patients with MC, 16% also had biopsy-proven celiac disease, a rate that is substantially higher than the 1% incidence of celiac disease in the general population. The presence of CD does not appear to modify the response to BSS in patients with MC, although larger cohorts are needed to answer this question more definitively. Sa1293 Endoscopic Appearance Correlates Poorly With Serology and Histology in Celiac Patients on a Gluten Free-Diet Jeffrey D. Jacobs, Matthew W. Stier, Carol E. Semrad, Sonia Kupfer Background: Endoscopic features of celiac disease (CD) have been reported at diagnosis. Less is known about the correlation of endoscopic appearance with serology and histology in CD patients on a gluten-free diet (GFD). In practice, some patients on GFD have abnormal endoscopic features with normal serology and histology, while others have normal endoscopic features with abnormal serology and histology. The causes and clinical consequences of these observations are unknown. The aim was to correlate endoscopic features with serology and histology in a large group of CD patients undergoing follow-up endoscopy. Methods: The electronic endoscopic database was queried for indications of "follow up of CD" or "established CD" between 8/2005-9/2014. Patients with anti-tissue transglutaminase (tTG) IgA or anti-deamidated gliadin peptide IgA (DGP) within 6 months of endoscopy were included. Endoscopic images were reviewed independently by two gastroenterologists blinded to clinical and pathology findings. Endoscopic features were classified as "normal", "mild" or "moderate to severe". Mild features included patchy scalloping or fissuring in the bulb or second part of the duodenum. Moderate to severe features included diffuse scalloping, fissuring, nodularity and/or atrophy. Inter-observer agreement was calculated using the unweighted Kappa statistic in R. Results: 200 CD patients (mean age 44 years; 80.4% female) with follow up endoscopies at a mean of 61 months from diagnosis were identified. 142 of 200 (71%) had serologies available within six months of endoscopy. Serology was abnormal in 52 of 142 (36.6%) patients. Histology was abnormal in 79 of 142 (55.6%) patients: 30% had both increased intraepithelial lymphocytes (IEL) and villous blunting, 18% had increased IEL alone, and 7% had villous blunting alone. Among these 142 patients, reviewers had moderate agreement on endoscopic findings (K=0.436). In cases with normal endoscopic features, 28.6% had both abnormal serology and histology. In cases with moder- ate-severe endoscopic features, 42.9% had both normal serology and histology. Conclusion: In CD patients undergoing follow up endoscopy, more than half of patients had abnormal histology. There was moderate correlation between gastroenterologists in classifying endo- scopic features. When reviewers agreed on normal endoscopic features, nearly a third of cases had abnormal serology and histology, underscoring the need for histological evaluation. Conversely, in cases with moderate-severe features on endoscopy, 43% of cases had normal S-284 AGA Abstracts serology and histology, suggesting a lag of macroscopic recovery or permanent alteration unrelated to inflammation. Endoscopic mucosal appearance is a poor predictor of inflamma- tion in follow-up for CD. This supports the need for biopsy and histological evaluation independent of endoscopic appearance. Sa1294 Lymphocytic Gastritis in Patients With Celiac Disease; Increasing Prevalence According to Age and Severity of Villous Atrophy Benjamin Lebwohl, Peter H. Green, Robert M. Genta Background: Lymphocytic gastritis (LG) is an uncommon histologic entity with varying symptoms and endoscopic appearances. An association with celiac disease has been reported, but the prevalence of LG in that population is unknown. We aimed to characterize the prevalence of LG in individuals with and without celiac disease, and to measure for a correlation between the degree of the villous atrophy and the prevalence of LG. Methods: We analyzed all concurrent gastric and duodenal biopsy specimens submitted to a national pathology laboratory during a six year period (January 2, 2008 through January 2, 2014). Patients with LG were classified as such if the gastric epithelium was infiltrated by at least 25 lymphocytes per 100 epithelial cells. We compared the prevalence of LG among those with normal duodenal histology (or non-specific duodenitis) and those with celiac disease, as defined by villous atrophy (Marsh 3). We performed multiple logistic regression to identify independent predictors of LG and included the following a priori variables in the multivariate model: age, gender, the presence of Helicobacter pylori , and duodenal histology. Results: Among patients who underwent concurrent gastric and duodenal biopsy (n=287,503), the mean age was 52 and the majority (67%) were female. LG was present in 818 (0.3%) patients, while H. pylori was present in 27,366 (10%). Duodenal intraepithelial lymphocytosis (DIL) was noted in 18,816 (7%) while celiac disease was present in 3,948 (1.4%). LG was slightly more common in patients with DIL (0.8%) and much more common in patients with celiac disease (7.3%, p<0.0001). Among patients with celiac disease, LG was more common among patients with total villous atrophy (9.7%) compared to those with partial or subtotal villous atrophy (5.0%, p<0.0001). On multivariate analysis, H. pylori colonization was not associated with LG (p=0.28). Independent predictors of LG were older age (OR for 60+ compared to 20-39: 1.83; 95%CI 1.51-2.21), the presence of DIL (OR 6.15; 95% CI 5.06-7.47), partial/subtotal villous atrophy (OR 37.66; 95% CI 30.16-47.03), and total villous atrophy (OR 78.57; 95% CI 65.37-94.44). Among patients with celiac disease ages ≥60 years, 10% had LG on gastric biopsy. Conclusion: While uncommon in the general population, LG is strongly associated with celiac disease, with increasing prevalence correlat- ing with more advanced villous atrophy and increasing age. Future investigations should measure the natural history of LG after celiac disease is treated with a gluten-free diet and assess the clinical significance of this prominent association. Sa1295 Reflux Symptoms in Patients With Celiac Disease Are Associated With Alkaline or Weakly Acid Reflux Fabio Nachman, María Inés Pinto Sanchez, Hwang Hui Jer, Edgardo Smecuol, Salma Salim, Horacio Vázquez, Ana F. Costa, Gabriela I. Longarini, María Laura Moreno, Sonia I. Niveloni, Roberto M. Mazure, Claudio R. Bilder, Andres Ditaranto, Premysl Bercik, Guido Iantorno, Eduardo Mauriño, Elena F. Verdu, Julio C. Bai Background/aims: Gastroesophageal reflux symptoms (RS) are frequently found in patients with active celiac disease (CD) and they revert with the gluten-free diet. Although previous studies have suggested that abnormal esophageal motility may underlie RS, motor abnormali- ties have not been consistently confirmed. Moreover, no studies have explored esophageal function in CD patients by combining functional tests such as esophageal pH-impedance and manometry. Aim: To evaluate esophageal function in newly diagnosed patients with CD compared to non-celiac patients with reflux symptoms and healthy volunteers. Methods: CD patients (defined with Marsh III, or greater, and positive tTG) were assessed at the time diagnosis independently of the presence of RS. A cohort of non-CD patients (negative tTG) with RS and healthy volunteers served as controls. The presence of RS was established by a sub-dimension in the Gastrointestinal Symptoms Rating Scale (GSRS), a cut off of ≥2 points was considered as moderate to severe RS. 24-hour ambulatory pH-impedance test was performed using Acutrac PHZ (Sierra Scientific, Los Angeles, CA). Data on upright and supine reflux episodes, DeMeester score, reflux episode duration and type of reflux were recorded. Manometry was performed using a water-perfused system (Dentsleeve Pty Ltd, Parkside, South Australia). Results: Thirteen patients with CD and 21 controls were enrolled in the study. Nine out of 13 CD patients and 18 out of 21 controls reported RS. Ten out of 13 CD patients had abnormal pH-impedance results vs. 16 out of 21 non-celiac with RS. All healthy controls had normal pH-impedance test. More episodes of weakly acid and non- acid reflux (p>0.001) and shorter episode duration ( p=0.03) were observed in CD vs. controls with RS. Notably, 75% of CD patients presented alkaline or weakly acid reflux (both at upright and supine positions), while all symptomatic non-celiac patients presented dominantly acid reflux. The DeMeester score was normal in CD patients and lower compared with non-CD patients with RS. CD patients with or without RS had a similar LES pressure as healthy controls while non-CD patients with RS had lower LES tone ( p=0.04). Conclusion: Differences in ambulatory pH-impedance were observed in CD patients compared to non-CD subjects, irrespective of the presence of RS. The type of reflux in CD patients was predominantly weakly acidic or non-acid, while in non-CD patients with RS the reflux was acidic. In CD patients, the LES pressure did not seem to play a major role in RS. The presence of mildly acidic or alkaline reflux in conjunction with the altered inter-cellular spaces previously demonstrated by us, may explain the presence of RS in CD.