to food antigens and/or environmental triggers, leads to the recruitment and infiltration of eosinophils in the esophageal mucosa. Aim: To characterize T cell responses in EoE in children. Methods: We enrolled 24 patients (mean age 14.1 years, range 10-17 years, 75% male), including 2 with newly diagnosed EoE, 11 with treated EoE (3 with active inflammation .15 eos/hpf, 8 in remission ,5 eos/hpf), 9 with normal endoscopy, 1 patient with mild reflux esophagitis ( ,5 eos/hpf) and 1 patient with Barrett's esophagus. We obtained 4 esophageal pinch biopsies at the time of clinically indicated endoscopy. For T cell isolation the biopsies were enzymatically digested. Single cell suspensions were analyzed by flow cytometry using specific cell markers for TCR αβ, CD3+, CD4+, and CD8+ T lymphocytes to phenotype the cells. A subset of cells were stimulated with PMA/ionomycin to measure intracellular cytokine potential (TNF- α and IFN-γ) via flow cytometry. Results: We extracted an average of 2.7x105 cells per patient (range 0.4to 6.5 x105). The average number of CD3 TCRαβ+ cells was 57% (range 3.9-82.9%) with 41% (range 11.3-66%) CD4+ and 44% (range 15.6-81.3%) CD8+. There was a trend towards a higher percentage of CD8+ T cells in EoE patients (n = 13) compared to the other groups (n = 11) that included normal controls and reflux patients (48% vs. 41%), although, this was not significant. The CD8+/ CD4+ ratio was also slightly higher in the EoE group compared to the other groups (1.87 vs. 1.25), although these results were not significant. In 20/24 patients (11 EoE and 11 others), on average, 29% of the CD3+ cells, 32.8% of the CD4+ cells and 41% of the CD8+ cells produced TNF- α when stimulated with PMA/ionomycin (n = 20). The average IFN- γ production (n = 20) was 19.5% of CD3+ cells, 15% of CD4+ cells and 37% of the CD8+ cells. There was no significant difference in TNF- α and IFN- γ production between EoE patients and the other groups. Conclusion: We demonstrated the feasibility of isolating CD3+ TCRαβ+ T lymphocytes from pediatric esophageal biopsies and the ability to phenotype and functionally test these lymphocytes. We have observed higher CD8+ T cells in EoE patients compared to other groups and a higher CD8+/CD4+ ratio in EoE patients compared to the other groups. We also observed lower TNF- α and IFN-γ production in EoE patients compared to the other groups. Further studies are needed to determine the role of these lymphocytes in the development of EoE. Su1831 Determination of the Molecular Phenotype of Esophageal Mucosal Inflammation in Children With Eosinophilic Esophagitis Using a 1-Hour Esophageal String Test (EST) Steven J. Ackerman, Amir Kagalwalla, Preeth Alumkal, Jian Du, Rachel Harris, Lindsay Hosford, Katie Amsden, Wendy Moore, Sophie A. Fillon, Joanne C. Masterson, Kelley Capocelli, Hector Melin-Aldana, Brian T. Maybruck, Sergei Ochkur, James J. Lee, Zhaoxing Pan, Glenn Furuta Background: Eosinophilic Esophagitis (EoE) is a chronic disease characterized by eosinophil- predominant esophageal inflammation for which endoscopy with biopsy is currently the standard method to assess mucosal inflammation and follow resolution with treatment. However, we recently reported that eosinophil secreted granule proteins (ESGPs) can be captured and assayed from esophageal secretions, and used to distinguish active EoE from EoE in remission, GERD and normal using a minimally invasive Esophageal String Test (EST) after leaving it in situ for 12-hours (1). Aims: Our objectives were to determine whether esophageal luminal biomarkers associated with esophageal eosinophilia could be captured by the EST in a shorter 1-hour sampling period and differentiate the activity of mucosal inflammation in children with EoE. Methods: Four hours before endoscopy with biopsy was performed, an EST was swallowed. After a 1-hour sampling period, the EST was removed and endoscopy was performed 3 hours later. ESGPs [major basic protein-1 S-485 AGA Abstracts (MBP1) and eosinophil peroxidase (EPX)], and Th2 cytokines (IL-4, IL-5, IL-13) and chemo- kines (Eotaxin-2, -3) were measured by ELISA in luminal effluents (secretions, cells) eluted from the ESTs and in extracts of esophageal biopsies obtained at the time of endoscopy (1). Results: Samples from 21 children (ages 9-18) with active EoE (n=5; 19-120 eos/HPF), treated EoE in remission (n=5; 0-9 eos/HPF), GERD (n=3; 0-2 eos/HPF) and normal esophagus (n= 8; 0-1 eos/HPF) were analyzed. EST measurements of MBP1 significantly differentiated between children with active EoE (8.07±3.6μg/ml), EoE in remission (0.59±0.17μg/ml), GERD (0.36±0.01μg/ml) and normal esophagus (0.04±0.02μg/ml) (all p ,0.01), while EPX distinguished between active EoE (0.97±0.48μg/ml), GERD (0.05±0.02μg/ml) and normal esophagus (0.07±0.02μg/ml) (all p ,0.01). EST measurements of eotaxin-2 differentiated between active EoE (756±356pg/ml) and treated EoE in remission (82±24pg/ml) (p ,0.05) and normal (99.7±15pg/ml, p=0.052), while Eotaxin-3 differentiated between active EoE (442±200pg/ml) and normal esophagus (2.5±2.5pg/ml) (p ,0.05). Th2 cytokines (IL-4, IL- 5, IL-13) were not detectable in 1-hour EST samples (all below sensitivity of the ELISAs). EST MBP1 (r=0.972, p, 0.001) and EPX (r=0.751, p ,0.001) biomarker levels correlated significantly with the peak number of eosinophils in biopsies, as did EST levels of Eotaxin-2 (r= 0.917, p,0.001) and Eotaxin-3 (r=0.917, p,0.001). Conclusions: Eosinophil-associated biomarkers that are captured by the EST in a clinically relevant time frame reflect mucosal inflammation and disease activity in EoE. The EST is a novel, time efficient, minimally invasive device for measuring esophageal eosinophilic inflammation in children with EoE. Reference: (1) Furuta GT et al. Gut. Aug. 15, 2012; PMID: 22895393. Su1832 Duration of Untreated Inflammation Represents the Main Risk Factor for Stricture Development in Eosinophilic Esophagitis Alain Schoepfer, Ekaterina Safroneeva, Christian Bussmann, Tanja Kuchen, Susanne Portmann, Alex Straumann Background: Eosinophilic Esophagitis (EoE) is a chronic-inflammatory condition presenting either as inflammatory-phenotype (IP), as stenosing-phenotype (SP) or as an overlap form. The IP is endoscopically characterized by whitish exudates, furrows and edema, whereas the SP shows signs reflecting tissue remodeling, such as rings and strictures. There is some evidence that SP develops over time, but the precise evolution of stricture formation during the disease course remains largely unknown. Aim: To correlate the prevalence of esophageal strictures with the duration of untreated disease and to evaluate risk factors for stricture formation. Methods: Retrospective analysis of the Swiss EoE Database (SEED), extended by a review of patients charts, endoscopy and pathology records. Strictures were assessed endoscopically at first diagnosis of EoE and defined as a narrowing leading to problems passing a standard adult upper endoscope (outer diameter 9mm). Diagnostic delay was defined as period from onset of EoE symptoms to diagnosis. Results: Two hundred EoE patients were analyzed (153 males, mean age at index visit 39 ± 15 years, all Caucasians). All patients were symptomatic; dysphagia was present in 94% and chest pain in 35% of patients. Allergies were identified in 132 (66%) of patients. An endoscopic bolus removal was performed in 56 patients (28%), whereof 28 prior and 28 at the time of EoE diagnosis. Median diagnostic delay was 6 years (IQR 2-12, range 0-36 years). We observed the following stricture prevalence at EoE diagnosis according to the following intervals of disease duration: 0-2 years (n=58) 17.2%, 3-5 years (n=39) 30.8%, 6-8 years (n=18) 38.9%, 9-11 years (n= 29) 37.9%, 12-14 years (n=12) 41.7%, 15-17 years (n=14) 64,3%, 18-20 (n=6) 66.7%, .20 years (n=24) 70.8% (p , 0.001, Cochrane-Armitage trend test). Logistic regression modeling detected a long duration of untreated disease (defined as diagnostic delay ≥ 7 years) to be strongly associated with the presence of strictures at the index visit (Odds Ratio 3.24, 95% confidence interval 1.75-5.97, p , 0.001). No association of strictures at the index visit was found with gender, presence of allergies, age at EoE diagnosis, positive family history of EoE, blood eosinophilia, elevated serum IgE, and peak eosinophil count at index histology. Conclusions: Strictures in EoE develop over time. The frequency of stricture formation is proportional to the duration of the untreated disease. These findings underscore the need to reduce the diagnostic delay in EoE and to control the underlying inflammation in order to prevent esophageal remodeling. Su1833 Eosinophilic Esophagitis Is a Progressive Fibrostenotic Disease: Insights From a Phenotypic Analysis Evan S. Dellon, Hannah P. Kim, Sarah McConville, David A. Rybnicek, John T. Woosley, Nicholas J. Shaheen Background: There is limited information about the natural history of eosinophilic esophagi- tis (EoE), and it is unknown whether specific phenotypes of EoE are the result of disease progression. Aim: To describe the clinical features of EoE patients with inflammatory, fibrostenotic, and mixed phenotypes, determine predictors of these phenotypes, and make inferences about the natural history of EoE. Methods: This was a retrospective study of the University of North Carolina EoE Clinicopathologic database from 2001-2011. Subjects with an incident diagnosis of EoE who met consensus guidelines were included. All had symptoms of esophageal dysfunction, ≥15 eos/hpf (hpf area=0.24 mm 2 ), and did not respond to a PPI trial. The phenotypes were defined as fibrostenotic if there were esophageal rings, narrowing, or strictures and no evidence of linear furrows or white plaques; as inflammatory if there were furrows, plaques, or a normal esophagus and no evidence of fibrostenotic changes; and as mixed if there were a combination of findings. Clinical and histologic features were compared between the three phenotypes. Multinominal logistic regression was performed to assess predictors of phenotype status. Results: Of 374 EoE cases, 134 (36%) were inflammatory, 163 (43%) were mixed, and 77 (21%) were fibrostenotic. Those with an inflammatory phenotype were more likely to be younger than those with mixed or fibrosteno- tic phenotypes (13 vs 29 vs 39 years, respectively; p ,0.001). Dysphagia was less likely in inflammatory compared to mixed or fibrostenotic (36% vs 77% vs 92%; p ,0.001), as was food impaction (15% vs 37% vs 39%) and esophageal dilation (0% vs 24% vs 47%; p ,0.001). Abdominal pain, vomiting, and failure-to-thrive were more common in inflammatory (p,0.001 for all), as were atopy and food allergies (p ,0.05). The mean symptom length prior to diagnosis was shorter for inflammatory compared to mixed and fibrostenotic (5 vs AGA Abstracts