S132 Abstracts of the XVII National Congress of Digestive Diseases / Digestive and Liver Disease 43S (2011) S115–S264 levels of IL-4, IL-17A and IFN-gamma compared to CD samples with no endoscopic recurrence and normal controls. In the same mucosal lesions, TNF-alpha and IL-5 content was higher than that seen in normal controls but unchanged as compared to areas with no endoscopic recurrence. Conclusions: Data indicate that, in contrast to “late” CD, early recurrence is characterized by accumulation of various Th cell subsets, which make distinct profiles of inflammatory cytokines. Notably, even in the absence of endo- scopic recurrence, CD mucosa is heavily infiltrated with cytokine-secreting inflammatory cells. OC.06.6 EXPRESSION OF GH/IGF-1 AXIS COMPOUNDS IN MODERATE- SEVERE CROHN’S DISEASE: ROLE OF MUCOSAL INFLAMMATION AND ANTI TNF-A TREATMENT M.L. Annunziata * , R. Caviglia, A. Micera, L.G. Papparella, M. Cicala Campus Bio Medico, Roma, Italy Background and aim: Growth failure in children, weight loss and catabolism in adults, are well known features of active CD. Increasing evidence shows that these features may be due to GH-resistance caused by persistent chronic inflammation. Aim of this study was to evaluate, in CD patients, a possi- ble peripheral GH-resistance (intestinal mucosa), and eventual modifications following anti TNF-a treatment. Material and methods: 9 adult patients with moderate-severe active CD, consecutively scheduled to receive three infliximab infusions, were studied. Biopsy specimens from normal appearing duodenal mucosa were collected, in the 2 weeks prior to the first and after the third infusion. Confocal and real time PCR analysis of IGF-1, phospho-STAT5 and SOCS-3 were carried out and normalised to GAP, 18S and H3 referring genes. Four dyspeptic patients represented our control group. Differences between CD patients and controls were analysed by ANOVA analysis. Results: In CD patients, a significant inverse correlation was found when comparing phospho-STAT5 levels with CDAI (r=-0.71; p=0.003). Confocal analysis showed a significant decrease in phospho-STAT5 immunoreactivity in sections from CD patients, compared to controls (p<0.001). After treatment, phospho-STAT5 levels significantly increased (p<0.001). Molecular data for phospho-STAT5 were in keeping with biochemical findings. At baseline, compared with controls, in CD patients, the PCR analysis of IGF-1, showed a trend towards a decrease, while SOCS-3 showed a trend towards an increase. Following treatment, SOCS-3 and IGF-1 target genes showed a trend toward a decrease and an increase, respectively. Nevertheless, following infliximab induction treatment, IGF-1, phospho-STAT5 and SOCS-3 levels did not reach those of the control group. Conclusions: The present study underline the effect of inflammation on the growth hormone axis not only at the level of hepatic metabolism, but also in the intestinal mucosa, since biologic agents able to inhibit pro-inflammatory cytokines, in particular TNF-a, the main inflammatory mediator, reverse growth hormone resistance. These therapeutic approaches could prevent the onset of those clinical conditions with a negative impact on child growth and quality of life of adults. OC.06.7 EFFECT OF MATRIX METALLOPROTEINASE (MMP)-3, -8, -10 AND -12 ON THE STRUCTURE AND FUNCTION OF ANTI-TUMOR NECROSIS FACTOR-ALPHA AGENTS IN INFLAMMATORY BOWEL DISEASE P. Biancheri * ,1 , A. Di Sabatino 1 , S.A. Snoek 2 , I. Joe-njoku 2 , L. Rovedatti 1 , N. Ahmad 2 , G.R. Corazza 1 , T.T. Macdonald 2 1 First Department of Medicine, Fondazione Irccs Policlinico S. Matteo, University of Pavia, Pavia, Italy; 2 Centre for Immunology and Infectious Disease, Blizard Institute of Cell and Molecular Science, Barts and The London School of Medicine and Dentistry, London, United Kingdom Background and aim: A considerable proportion of inflammatory bowel dis- ease (IBD) patients fail to respond to anti-tumor necrosis factor (TNF)-alpha treatment. Several matrix metalloproteinases (MMPs) can cleave human IgG proximally to the hinge region. Anti-TNF-alpha agents are IgG or contain part of the IgG sequence. We therefore investigated the cleavage of anti-TNF-alpha agents by MMPs that are up-regulated in IBD inflamed mucosa, namely MMP-3, -8, 10 and -12. Material and methods: Activated recombinant human MMP-3, -8, -10 or -12 were incubated with infliximab, adalimumab, certolizumab pegol or etaner- cept. IgG Fc portion or Ig kappa light chains were detected by immunoblotting on the supernatants collected at 0, 3, 6, 10 and 24h. The ability of MMP-treated antibodies to neutralise TNF-alpha was tested using a luciferase reporter cell line. Mucosal homogenates from inflamed areas of 6 IBD patients and from normal gut of 6 control subjects were depleted of endogenous IgG and incubated for 24h with the four drugs, then immunoblotted for human IgG Fc portion or kappa light chains. Results: Certolizumab pegol was not cleaved any of the MMPs tested. In- fliximab and adalimumab were degraded only by MMP-3 and -12, releasing 32kDa Fc monomers and F(ab)2. Etanercept was cleaved by MMP-3, -10 and -12, although Fc monomers were found only after MMP-3 and -12 digestion. All the four drugs inhibited luciferase production in TNF-alpha-stimulated HeLa 57A cells. When the anti-TNF-alpha agents were pre-incubated with ac- tivated MMP-3, -10 or -12, only etanercept lost its ability to inhibit luciferase production. None of the anti-TNF-alpha agents were degraded by the control mucosa, whereas etanercept was cleaved by IBD tissue. Conclusions: Adalimumab and infliximab are functioning as F(ab)2 frag- ments after degradation by MMP-3 and -12, whereas etanercept is unable to neutralise soluble TNF-alpha after MMP-cleavage. These findings provide a possible explanation for the clinical inefficacy of etanercept in IBD patients. OC.07.1 MODULATE COLITIS SURVEILLANCE: FOXE1 AND SYNE1 HYPERMETHYLATION IN COLITIS ASSOCIATED COLORECTAL CANCER (CAC) C. Papadia * ,1 , J. Louwagie 2 , P. Del Rio 7 , M. Grooteclaes 2 , A. Coruzzi 1 , C. Montana 1 , A. Franzè 1 , M. Novelli 8 , C. Bordi 3 , W. Bryan 4 , W. Atkin 5 , A. Forbes 6 1 Gastroenterology, University of Parma, Parma, Italy; 2 Research and Development, Mdxhealth S.A., Liege, Belgium; 3 Dept. of Pathology, University of Parma, Parma, Italy; 4 Dept. of Pathology, University of Oxford, Oxford, United Kingdom; 5 Surgery and Cancer, Imperial College, London, United Kingdom; 6 Gastroenterology, University College Hospital, London, United Kingdom; 7 Dept. of Surgery, Parma University Hospital, Parma, Italy; 8 Dept. of Pathology, University College London, London, United Kingdom Background and aim: Surveillance for colorectal cancer is necessary for patients with inflammatory bowel disease (IBD). Colitis-associated colorectal cancer (CAC) affects individuals at a younger age than the general popula- tion. Colonoscopy is the gold standard screening method for longstanding Ulcerative Colitis (UC) and Crohn’s colitis (CD). Changes to the surveillance intervals have been made in the light of recent data demonstrating that endoscopic appearance is an important predictor of future dysplasia or cancer development. Clinical standards in screening outcomes are far from optimal and a methylation markers panel may be greatly useful. Material and methods: Using re-expression profiles of colon cancer cell lines, candidate genes were identified and the most promising markers were tested on tissue using the Base5 methylation profiling platform. Promoter sequences were linked with gene expression to identify epigenetically silenced genes. Marker candidates identified by re-expression were screened using methylation specific PCR assays to assess the methylation status of 2 gene promoters (FOXE1, SYNE1) in 127 formalin-fixed paraffin-embedded (FFPE) tissue samples collected from colonoscopy surveillance in (n=97) longstanding IBD patients and thirty (n=30) healthy controls. Samples included colorectal adenocarcinomas, on background of IBD, of various stages (n=26); flat IBD dysplastic lesions (n=29); dysplastic adenomas arising on background of UC (n=4); samples from patients with no evidence of dysplasia or cancer but longstanding IBD (n=38) and healthy controls (n=30). Results: The analytical cut-off was set to give 100% specificity (based on the thirteen healthy controls). The combination of the tested markers gave