AGA Abstracts Sa1891 A SURVEY OF DISCLOSURE PREFERENCES IN PATIENTS WITH INFLAMMATORY BOWEL DISEASE Ling Guo, Andrew Canakis, Francis A. Farraye Introduction: Enacted stigma is the social discrimination towards those in the stigmatized group and is associated with decreased QoL, increased rates of depression and anxiety in patients with IBD. Disease disclosure is complex but has been found to have positive benefits. Recent data from our group demonstrated that disclosure of IBD status led to decreased rates of enacted stigma from others (J Health Commun. 2018;23:91-97). This study aims to examine the habits and preferences of disclosure in patients with IBD. Methods: English- speaking adults with an established diagnosis of IBD (age ≥ 18) seen at Boston Medical Center's IBD practice were enrolled in October 2018. Patients completed a 2- page question- naire detailing their disease severity, disclosure habits and preferences. We assessed for work absenteeism and medication compliance. Depression and anxiety were evaluated using the PHQ-2 scale and the Generalized Anxiety Disorder-7 scale, respectively. Demographic information, remission status and type of IBD medication was recorded. Results: Forty- eight patients completed the questionnaire; 46% were female, 54% had Crohn's Disease, 42% had Ulcerative Colitis. The mean age was 36 (range 19-73). The mean HBI and SCCAI scores were 2.19 and 2.06, respectively. Sixty percent were in remission based on their last endoscopy and 52% felt their IBD was well controlled. Out of the 42 people who were on IBD medications, 13 (31%) reported some difficulty with medication compliance and of the 17 people who reported work absenteeism, 14 of them missed work as a result of their IBD. Fifteen percent of patients screened positive for depression. Most respondents were happy with their decision to disclose with only 8% noting discontent. All patients had disclosed their IBD to others, with family and close friends being the most frequently disclosed to, followed by significant others, and colleagues/classmates. The most common reasons for nondisclosure are listed in table 1. When asked about what resources would make the disclosure process easier, 40% responded that research showing disclosure can decrease stigma would be most helpful, followed by a sample disclosure script, more informa- tion about IBD disease course and treatments, small group sessions, and finally the use of a decision aid. Conclusions: All of the respondents in our study noted some disclosure of their IBD status with the majority documenting a positive experience. The most common reason for non- disclosure was privacy, followed by other people's disbelief, repugnance of the disease and outward stigma, which encompasses elements of enacted, perceived and internalized stigma. The majority of respondents noted that resources would be helpful in easing the disclosure process. Future studies should examine the connection between disease disclosure and decreased stigma. Reasons for non-disclosure Don't think everyone needs to know 41% Fear that people won't believe you if you 15% don't have symptoms they can see Fear of being stigmatized 11% You find stool, (bloody) diarrhea, 11% ostomy bags gross or dirty You think others will find stool, (bloody) diarrhea, ostomy bags gross or 11% dirty Fear of being treated differently 4% Bad experiences from telling other peo- 4% ple Having IBD or the symptoms is a result 4% of something you did Sa1892 NUTRITIONAL STATUS CORRELATES WITH QUALITY OF LIFE IN ACTIVE CROHN'S DISEASE Cristina Tocia, Luana Alexandrescu, Eugen Dumitru INTRODUCTION Hospitalized patients with Crohn's disease (CD) have an impaired quality of life (QoL) and can develop severe nutritional deficits. SIBDQ is a rapid and reliable measure of QoL and an important patient-reported outcome. Nutritional screening tools (NST) with the most utility in this patients are still unknown. Aim of this study was to assess the NST which accurately reflect the disease activity and the impact on QoL in hospitalized patients with CD. MATERIAL AND METHOD The prospective study included 89 patients with active CD admitted to our hospital. Disease activity was asessed by CDAI score, QoL by SIBDQ and nutritional status by MUST score, NRS 2002, O-PNI, CONUT and albumin levels. Inclusion criteria were: hospitalized CD patients with CDAI > 150, age more than 16 years old. Exclusion criteria: disease in clinical remission (CDAI < 150). Non- parametric statistics were used to compare groups of patients and correlation was analysed using Spearman rank coefficient (r). Two-sided p < 0.05 was considered statistically signifi- cant. RESULTS Assesment of disease activity by CDAI Moderate disease was present in 52 (58%) patients, severe disease in 21 (24%) patients and mild disease and 16 (18%) patients. Correlations CDAI was closely correlated with albumin levels (r = - 0.70, p < 0.05), O-PNI (r = - 0.60, p < 0.05), MUST (r = 0.51, p < 0.05) and NRS 2002 (r = 0.42, p < 0.05). An insignificant correlation between CDAI and CONUT was found (r = 0.18, p > 0.05). SIBDQ was correlated with albumin (r = 0.68, p < 0.05), O-PNI (r = - 0.61, p<0.05), NRS 2002 (r = 0.50, p < 0.05) and MUST (r = - 0.44, p < 0.05). An insignificant correlation between SIBDQ and CONUT was found (r = - 0.14, p > 0.05). CONCLUSIONS Albumin, NRS2002, MUST and O-PNI are useful NST and reflect the impact of CD activity and also the impact of the disease on QoL assesed by SIBDQ. S-444 AGA Abstracts Sa1893 A SINGLE CELL SURVEY OF LARGE INTESTINE IN MICROSCOPIC COLITIS Stefan Halvorsen, David Morgan, Kaia C. Miller, Yueming Cao, Jessica McGoldrick, Kyle Staller, Daniel C. Chung, Slim Sassi, Hamed Khalili Background: A significant challenge in studying microscopic colitis (MC) has been its remarkably similar mucosal and histologic appearance to that of healthy colonic mucosa. Here, we leverage single-cell RNA-seq (scRNA-seq) technology to characterize the heterogene- ity in epithelial and immune cells derived from mucosal biopsies of patients with MC. Methods: We collected random colon mucosal biopsies from 5 MC patients and 10 healthy controls. Biopsies were enzymatically digested into single cell suspensions. scRNA-seq was applied using an emulsion-based method, InDrops. For each patient, we constructed 2 libraries containing 2,000-3,000 cells. Libraries were sequenced using the Illumina HiSeq2000 platform. Sequencing reads were mapped to the human genome (GRCh38) using the STAR aligner. Unique molecular identifiers (UMIs), a 6 random sequence used to tag individual molecules prior to amplification enabling digital mRNA quantification, were collapsed and called from the aligned reads using the DropSeqTools DigitalExpression program. The UMI matrix was then inputted into the Seurat R package for downstream analysis. The UMI counts were normalized and scaled, followed by principal component analysis (PCA) based clustering. T-distributed Stochastic Neighbor Embedding (tSNE), a machine learning algorithm used for visualization of high dimensional data, was performed using the top 20 principal components as input. Results: After quality control, 20,274 cells derived from 15 patients were included in our analyses. The mean number of genes identified per cell was 736 (standard deviation = 493). We identified two clusters of cells that were unique to patients with MC (Figure 1). One cluster was characterized by high expression of KRT18, KRT20, AQP8, CXCL11, ISG15, NOS2, IFIT3, and HLA genes demonstrating an epithelial cell population (Figure 2a) while the other cluster was characterized by high expression of CD3D, CD8A, NKG7, PRF1, and CD7 genes, representing a population of CD8+ cytotoxic T lymphocytes (Figure 2b). Gene set enrichment analyses demonstrated a significant overlap between gene sets in the unique epithelial cell population and genes induced in response to INF-gamma (q = 1.08E-51), graft versus host disease (q = 7.63E- 34), and antigen processing and presentation (q = 7.45E-26). Conclusion: We present preliminary data on an unbiased transcriptional profiling at the single-cell resolution for MC. Our data demonstrate that primary immune cells involved in the pathogenesis of MC appear to be antigen-specific CD8+ cytotoxic T cells. We also describe a unique population of interferon-responsive colonic epithelial cells with increase expression of MHC class I and II genes. Our approach holds significant potential for uncovering fundamental aspects of MC. Sa1894 ENTERIC GLIAL MHC II CONTRIBUTES TO T-LYMPHOCYTE ACTIVATION IN LPS STIMULATION Aaron K. Chow, Brian D. Gulbransen Introduction: Enteric glia contribute to the regulation of physiological gut reflexes and to neuroinflammation in disease. How glia contribute to gut inflammation through immune cells is poorly understood. Enteric glia express major histocompatibility class II (MHC II) during various forms of intestinal inflammation in mice and humans. Therefore, we hypothe- size that antigen presentation by enteric glia contributes to a pro-inflammatory environment during ENS injury. Methods: We tested our hypothesis by generating mice with a targeted ablation of MHC II in glia (Sox10CreERT2::IABfl/fl). Animals were injected via intraperitoneal injection with 1μg interferon-gamma (IFNg) and 300μg/kg lipopolysaccharide (LPS) or saline control, and harvested 16 hours post-injection. Our genetic knockout was validated by