Aravinth U. Murugananthan

A mechanistic role for leptin in human dendritic cell migration: differences between ileum and colon in health and Crohn's disease

Dendritic cells (DC) migrate to lymph nodes on expression of C-C motif chemokine receptor 7 (CCR7) and control immune activity. Leptin, an immunomodulatory adipokine, functions via leptin receptors, signaling via the long isoform of receptor, LepRb. Leptin promotes DC maturation and increases CCR7 expression on blood DC. Increased mesenteric fat and leptin occur early in Crohn’s disease (CD), suggesting leptin-mediated change in intestinal CCR7 expression on DC as a pro-inflammatory mechanism. We have demonstrated CCR7 expression and capacity to migrate to its ligand macrophage inflammatory protein 3β in normal human ileal DC but not colonic or blood DC. In CD, functional CCR7 was expressed on DC from all sites. Only DC populations containing CCR7-expressing cells produced LepRb; in vitro exposure to leptin also increased expression of functional CCR7 in intestinal DC in a dose-dependent manner. In conclusion, leptin may regulate DC migration from gut, in homeostatic and inflammatory conditions, providing a link between mesenteric obesity and inflammation.

Chemokine (C-C Motif) Receptor 2 Mediates Dendritic Cell Recruitment to the Human Colon but Is Not Responsible for Differences Observed in Dendritic Cell Subsets, Phenotype, and Function Between the Proximal and Distal Colon

Background & Aims Most knowledge about gastrointestinal (GI)-tract dendritic cells (DC) relies on murine studies where CD103+ DC specialize in generating immune tolerance with the functionality of CD11b+/− subsets being unclear. Information about human GI-DC is scarce, especially regarding regional specifications. Here, we characterized human DC properties throughout the human colon. Methods Paired proximal (right/ascending) and distal (left/descending) human colonic biopsies from 95 healthy subjects were taken; DC were assessed by flow cytometry and microbiota composition assessed by 16S rRNA gene sequencing. Results Colonic DC identified were myeloid (mDC, CD11c+CD123−) and further divided based on CD103 and SIRPα (human analog of murine CD11b) expression. CD103-SIRPα+ DC were the major population and with CD103+SIRPα+ DC were CD1c+ILT3+CCR2+ (although CCR2 was not expressed on all CD103+SIRPα+ DC). CD103+SIRPα- DC constituted a minor subset that were CD141+ILT3−CCR2−. Proximal colon samples had higher total DC counts and fewer CD103+SIRPα+ cells. Proximal colon DC were more mature than distal DC with higher stimulatory capacity for CD4+CD45RA+ T-cells. However, DC and DC-invoked T-cell expression of mucosal homing markers (β7, CCR9) was lower for proximal DC. CCR2 was expressed on circulating CD1c+, but not CD141+ mDC, and mediated DC recruitment by colonic culture supernatants in transwell assays. Proximal colon DC produced higher levels of cytokines. Mucosal microbiota profiling showed a lower microbiota load in the proximal colon, but with no differences in microbiota composition between compartments. Conclusions Proximal colonic DC subsets differ from those in distal colon and are more mature. Targeted immunotherapy using DC in T-cell mediated GI tract inflammation may therefore need to reflect this immune compartmentalization.

What role do bacteria play in persisting fistula formation in idiopathic and Crohn's anal fistula?

The aetiology of Crohns disease‐related anal fistula remains obscure. Microbiological, genetic and immunological factors are thought to play a role but are not well understood. The microbiota within anal fistula tracts has never been examined using molecular techniques. The present study aimed to characterize the microbiota in the tracts of patients with Crohns and idiopathic anal fistula.

T-cell proliferation and forkhead box P3 expression in human T cells are dependent on T-cell density: physics of a confined space?

T-cell proliferation rates in vitro depend on factors including initial T-cell number, dose of stimulus, culture time, and available physical space. The role of forkhead box P3 (FoxP3) in the identification of T cells with a regulatory phenotype remains controversial in humans. Through 5-carboxyfluorescein diacetate succinimidyl ester labeling of human T cells and subsequent culture of different numbers of T cells and antigen-presenting cells (APC), we studied proliferative T-cell responses and FoxP3 expression in divided T cells. T-cell proliferation rates depended on initial T-cell/APC numbers. Proliferation rates decreased when high initial T-cell numbers were increased. FoxP3 expression was expressed exclusively in virtually all divided T cells cultured at high T-cell densities, irrespective of their CD4 nature or cytokine content, and was coexpressed with T-bet. However, when T cells were cultured on larger surfaces or at lower initial numbers, FoxP3 expression was not induced in divided T cells, even when most of the cells had undergone cell division. FoxP3(+) T cells generated at high cell densities did not elicit a suppressive phenotype and FoxP3 expression was subsequently lost in time when the stimulus was removed. Therefore, caution should be observed in the use of FoxP3 expression to identify regulatory T cells in humans because its expression may be only a consequence of activation status in a restricted environment.

PMO-231 Ileal and colonic mucosal dendritic cell cytokine profiles differ at rest and after in vitro bacteria and pro-biotic challenge in postoperative Crohn's disease patients

Introduction Postoperative Crohns disease (CD) recurrence predominantly affects the ileal mucosa at the ileo-colonic anastomosis with the colonic side often spared. Altered immune responses to bacterial flora are thought to be a driving force in the pathogenesis of CD recurrence. Gut dendritic cells (DC) are key in the initiation of immune response, through cytokine production, when stimulated with bacterial antigens. We postulate that differences between ileal and colonic DC resting characteristics and functional responses may be responsible for the propensity of recurrence to occur at the ileal aspect of the anastomosis. We aimed to assess ongoing intracellular cytokine production in DC from ileal and colonic postoperative CD mucosa and assess their functional response to bacterial stimulation and modulation with probiotics. Methods Paired ileal and colonic biopsies were taken from postoperative CD patients at colonoscopy (n=11). Lamina propria mononuclear cells were collected after collagenase digestion. DC intracellular cytokine responses (IL-2, IL-6, IL-17a, TGFβ and INFγ) were assessed in basal conditions and after culture with LPS and two probiotic bacterial strains Bifidobacterium Longum; Lactobacillus Casei (B longum and L casei) using multi-colour flow cytometry. Results Unstimulated ileal DC showed higher levels of ongoing intracellular production of pro-inflammatory cytokines than unstimulated colonic DC: IL6 (34.21±12.80 vs 10.47±3.574 cells/υl [mean±SEM], p=0.037), IL17a (24.62±12.38 vs 14.94±9.865 cells/υl, p=0.05, n=5) and TGFβ (74.12±17.96 vs 32±16.27 cells/υl, p=0.031). Incubation with LPS resulted in higher DC intracellular cytokine levels of INFγ in ileal derived DC with a borderline p value (27.49±12.61 vs 0.39±0.391 cells/υl p=0.06) but not colonic derived DCs (19.55±10.12 vs 12.40±7.039, p=0.6). L casei incubation, however, led to a larger decrease in ongoing TGFβ (−42.35±16.02 vs 4.42±11.46 cells/υl p=0.023) and INFγ (−14.76 ±7.196 vs 20.33±10.16 cells/υl-, p=0.05) DC cytokine production in colonic tissue compared with ileal. Conclusion Ileal mucosa DC demonstrate a cytokine profile implicating a Th17 response compared with colonic mucosa. Upon bacterial stimulation with LPS ileal mucosa demonstrate increased INFγ DC production compared with unstimulated DC. These results suggest a role of for a Th1/Th17 response in driving post-operative CD recurrence. The probiotics L casei and B longum failed to show significant effects in modulation of intracellular cytokine production in ileal DC. Competing interests None declared.

PMO-199 Comparative study of sample adequacy of 25G vs 22G needle in endoscopic ultrasound (EUS) guided fine needle aspirate (FNA) of solid lesions

Introduction Optimal needle size in achieving greatest diagnostic yield from EUS- guided FNA remains unclear. Aim We prospectively compared sample adequacy and safety of FNA of solid lesions between 25G and 22G (CookTM) needle at two tertiary centres. Methods Prospective data from two sites was collected between November 2008 and November 2011. A single operator alternated on a case-by-case basis between a 25G and 22G needle. A cytopathologist was present to assess adequacy of sample. The operator could switch needle size if required. Results 152 patients undergoing 165 FNA were analysed (42M/30F, mean age 59). 76 patients had FNA with a 22 F needle and 76 with the 25F needle. Indications for EUS and FNA were pancreatic lesions 43%, lymph node enlargement 28%, biliary tract lesions 16%, submucosal lesion 8% and adrenal mass 1% and others 4%. Overall sample adequacy was 83.03% Adequacy per needle was 86.7% (22G) vs 79.2% (25G), p=0.22 Fischers Exact test. The number of passes used in successful FNA was higher with use of the 25G needle compared with the 22G needle. (2.42±0.11 SEM vs 1.962±0.15 SEM, p=0.015, t-test). In particular the use of a 25G needle had a higher number of passes in pancreatic lesions compared with the 22G needle (2.58±0.16 SEM vs 1.94±0.14, p=0.004, t-test). There was no difference in adequacy between the needle sizes for each type of lesion sampled (Abstract PMO-199 table 1). Two needle exchanges (25G to a 22G) occurred. One complication of local site bleeding occurred (22G) that settled during the test.Abstract PMO-199 Table 1 Lesion site 22G 25G Fischers exact test Adequate sample Inadequate sample Adequate sample Inadequate sample Lymph node 19 4 20 3 NS Biliary tract lesion 9 2 11 4 NS Pancreatic lesion 33 3 33 2 NS Submucosal lesion 4 2 2 5 NS Conclusion We show no difference in sample adequacy between the two needle sizes. Use of a 25G results is associated with a higher number of passes in pancreatic FNA. Both needle sizes appear safe. Operator choice and ease of passage of needle into anatomical location may also influence choice of needle. Competing interests None declared.

PWE-211 Longitudinal surveillance of submucosal tumours by endoscopic ultrasound: a single operator experience

Introduction Endoscopic ultrasound (EUS) provides increasingly improved assessment of submucosal tumours (SMT). The low yield of fine needle aspiration and the benign nature of the majority of lesions leave the optimal management for smaller, asymptomatic lesions unclear as further invasive procedures or surgery may be avoidable. Longitudinal studies characterising interval change of SMT are lacking and may provide information useful in optimising management. We report the experience of a single operator in the EUS surveillance of SMTs. Methods Patient cases were reviewed at two tertiary referral hospitals and one private hospital for patients who had serial EUS of the same lesion. EUS was performed by a single operator (RYC). For patients who had more than two EUS studies, details of the first and last were examined. Site, maximal diameter and layer of involvement were recorded. Paired data were analysed using a paired t-test with tests for correlation. Results 73 patients with SMT had at least two EUS procedures between February 2002 and October 2011. Lesions were found in the oesophagus (14), stomach (51) and duodenum (8) with involvement of the submucosa (40), deep submucosa/muscularis propria (4) and muscularis propria (29). The range between first and last EUS was 4–80 months. The lesions varied between 4 and 50 mm with a mean maximal diameter of 15.13 mm (95% CI 12.78 to 17.49) for the first EUS and 15.73 mm (95% CI 13.36 to 18.09) for the second EUS. Paired t-test analysis between the first and last measurements show that there was no significant difference (p=0.2321), with good correlation and effective pairing (r=0.9128, p<0.0001). Conclusion In our study of submucosal tumours smaller than 50 mm, we showed no significant change in size on surveillance EUS. Further longitudinal studies are needed to determine optimal surveillance regimen for SMTs. Competing interests None declared.

PTU-238 Prep, no prep or more prep? a prospective randomised study comparing two bowel preparation regimes with no preparation on quality of capsule endoscopy

Introduction Capsule endoscopy (CE) is a widely used method for evaluation of the small bowel. However it does have limitations; visualisation of the small bowel mucosa is often impaired due to the presence of food residue, air bubbles and bile pigments.1 The effect of bowel preparation on improving visualisation of the small bowel varies2 and is inconvenient for patients.3 We aimed to prospectively evaluate the effects of two different bowel preparations on visualisation of the small bowel and on overall diagnostic yield compared with standard dietary changes. Methods 51 patients (26 male/25 female; mean age 60.7 years) were randomised into three groups using the sealed envelope technique. Indications for CE were iron deficiency anaemia, obscure GI bleeding (occult and overt) and anaemia. Group 1 (n=19): Clear fluid day before procedure. Overnight fast. Group 2 (n=12): Clear fluid day before procedure. 2L PEG in afternoon of day prior to procedure. Overnight fast. Group 3 (n=20): Clear fluid day before procedure. 1L PEG and 1 sachet Picoprep in afternoon of day prior to procedure. Overnight fast. CE were viewed by a single blinded examiner and adequacy of bowel preparation according to three categories (>80% visualisation; 50%–80% visualisation). Results Conclusion Our findings are in keeping with a recent meta-analysis which has shown no difference in CE completion rates, GTT and SBTT with purgative preparation.4 Our study shows a trend towards better caecal completion rates with bowel preparation involving PEG and Picoprep, but these results did not reach statistical significance. Overall diagnostic yield was similar in all three groups. Liquid diet, in combination with fasting, prior to CE is generally better tolerated by patients3 and our findings would support this as adequate preparation for CE. Mean age Completion rate (%) Yield (%) Good SB views (%) Mean GTT ± SEM (min) Mean SBTT ± SEM (min) Gp 1 63.6 79 42.1 100 35.9±11.19 254.8±24.83 Gp 2 57 83.3 41.6 81.2 87.5±47.79 239.3±45.7 Gp 3 60 90 35 79 74.8±27.06 211.5±24.14 p Value NS NS NS NS Competing interests None declared. References 1. Park SC, Keum B, Seo YS, et al. Effect of bowel preparation with polyethylene glycol on quality of capsule endoscopy. Dig Dis Sci 2011;56:1769–75. 2. Dai N, Gubler C, Hengstler P, et al. Improved capsule endoscopy after bowel preparation. Gastrointest Endosc 2005;61:28–35. 3. Pons Beltrán V, González Suárez B, González Asanza C, et al. Evaluation of different bowel preparations for small bowel capsule endoscopy: a prospective, randomised, controlled study. Dig Dis Sci 2011;56:2900–5. 4. Rokkas T, Papaxoinis K, Triantafyllou K, et al. Does purgative preparation influence the diagnostic yield of small bowel video capsule endoscopy? : a meta-analysis. Am J Gastroenterol 2009;104:219–27.

Tu1552 Comparative Study of Sample Adequacy of 25G vs. 22G Needle in Endoscopic Ultrasound (EUS) Guided Fine Needle Aspirate (FNA) of Solid Lesions

budding sign, 5) irregular narrowing (2). 88% (n 58) patients underwent endoscopic treatment and 12% (n 9) patients surgery. Diagnostic accuracy of EUS for T staging of the EGC was determined by comparing preoperative EUS with subsequent postoperative histopathology findings. Results: Depending on the growth of EGC within the mucosal layer diagnostic accuracy of EUS was 93%, within submucosal layer 66%, within muscular layer 66%. The overall diagnostic accuracy of EUS in preoperative determination of EGC depth of invasion was 75%. Depending on the type of the EGC accuracy of EUS was varying: type 0-Ips 78%, type 0-IIa 87,5%, type 0-IIa/c 78%, type 0-IIb 0, type 0-IIc 75%, type 0-IIc/IIa 75%. Predominant location and distribution of the EGC in the stomach were in the antrum 50%, corpus (lower part 16,5%, middle part 9%, upper part 9%), angle of stomach 10,6%, cardia 5%. The rates for overstaging and understaging were 20% and 5%, respectively. Conclusion: EUS is a useful diagnostic method for preoperative staging of EGC for T criteria. However, EUS tended to overstage T criteria, and misdiagnosis increases with depth of tumor invasion. The least accurate EUS was at type 0-IIc and type 0-IIc/IIa. The main reasons of misdiagnosis were biopsy effect, wrong interpretation of the submucosal layer, peristaltic movements and perifocal inflammatory changes.

Mo1763 Clinical Risk Factors for Crohn's Disease Postoperative Recurrence are Reflected in Alterations in Mucosally Adherent Microbiota at Surgical Resection

Introduction Clinical risk factors for Crohn9s disease (CD) recurrence after ileo-caecal resection (ICR) include smoking status, perforating disease and >1 surgical resection. The underlying mechanisms contributing to clinical risk are unknown. We aimed to study the relationship between risk factors and gut microbiota. Methods Samples of macroscopically inflamed and non-inflamed small bowel from patients undergoing surgical resection for CD were analysed. Samples were snap frozen in liquid nitrogen. Cryosections were cut and the frozen sections were hybridised with oligonucleotide probes targeting the microbial 16S rRNA of total bacteria, Escherichia coli , Bacteroides-Prevotella, Faecalibacterium prausnitzii , Clostrium coccoides - Eubacterium rectale and bifidobacteria. The hybridised mucosa associated microbiota (MAM) were identified and quantified. Patients with ≥1 risk factor were classified as high risk for disease recurrence. Results Fifteen patients underwent ICR (10 female); 9 were high risk (6 smokers, 4 fistulating disease and 2 recurrent resection- 3 patients had multiple risk factors). Faecalibacterium prausnitzii numbers in inflamed operative samples were lower in smokers compared with non-smokers (p=0.036). High-risk patients had lower numbers of bifidobacteria in both inflamed (p=0.006) and non-inflamed (p=0.01) operative samples compared with low risk patients. Conclusion The risk of post-operative CD recurrence may be predetermined at a pre-operative stage due to dysbiosis. The role of MAM as a tool to stratify risk requires further study. Drugs that modulate MAM may, in future, play a role in reducing post-operative recurrence. Competing interests None declared.