AGA Abstracts least 220 were randomized to naltrexone (4.5 mg daily) or placebo by mouth daily for 12- weeks. Subsequently, those who had received placebo were treated with naltrexone and those randomized to naltrexone continued on naltrexone for an additional 12 weeks to assess safety and maintenance of response for extended duration. Colonoscopies with biopsies were performed at baseline, weeks-12, and 24 to determine endoscopic and histologic scores by persons blinded to the treatments. The primary outcomes included the CDAI, endoscopic, and histologic inflammatory scores. Other outcomes included quality of life (QOL) according to the IBDQ and SF-36 surveys, and laboratory tests for safety. Results: Forty patients were enrolled and randomized into the study. Baseline CDAI scores (346±16.4) indicated moderate to severe disease. CDAI scores decreased significantly from baseline (p<0.001) in those treated with naltrexone, but not in placebo treated controls. Eighty-two percent of the naltrexone-treated subjects had at least a 70-point drop in the CDAI score indicative of a response and 45% achieved clinical remission (CDAI < 150). Endoscopy and histology inflammatory scores improved in naltrexone treated subjects (p=0.0019) with evidence of decreased mucosal inflammation, and restored crypt architecture. No endoscopic or histolog- ical change was found in placebo treated controls. No significant changes were noted in laboratory parameters or QOL surveys. Patients in the placebo group reported increased fatigue compared to naltrexone treated subjects (p=0.04) and two naltrexone-treated subjects had transient mild elevation in liver transaminases which resolved without interruption of therapy. Conclusion: Naltrexone induces mucosal healing and decreases CDAI scores in subjects with moderate to severe Crohn's disease with minimal side effects recorded. Strategies to alter the endogenous opioid system provide promise for a novel treatment of inflammatory bowel disease in the future. (Supported by BMRP IBD-0180R and NIH DK073614). 647 Chemokine Receptor Antagonist CCX282-B (Traficet-En™) Maintained Remission of Crohn's Disease in PROTECT-1 Study Satish Keshav, Daniel Johnson, Thomas Schall, Pirow Bekker Introduction: CCX282-B (Traficet-EN), an orally bioavailable, CCR9-specific chemokine receptor antagonist showed efficacy in a 12-week Induction study. The results from a 36- week Maintenance study are presented here. Methods: PROTECT-1 is a randomized, double- blind, placebo-controlled trial in 436 subjects with moderate to severe Crohn's disease (Crohn's Disease Activity Index [CDAI] 250-450 and C-reactive protein [CRP] >7.5 mg/L). After a 12-week Induction and 4-week Open-label phase, CDAI ≥70-point responders were re-randomized to receive placebo or CCX282-B 250 mg twice daily for 36 weeks. Endpoints included CDAI response and remission at 36 weeks. Results: 241 subjects were randomized, 146 to CCX282-B and 95 to placebo. The mean Maintenance phase baseline CDAI was 138 (± 73) and median CRP 13 mg/L. The proportions of subjects in remission (CDAI ≤150) over 36 weeks are shown in the figure. CCX282-B maintained remission rate at ~50%, whereas it decreased progressively from 50% to 31% on placebo. At 36 weeks, 47% of subjects on CCX282-B were in remission vs. 31% on placebo (p = 0.01). Furthermore, at 36 weeks, 41% of subjects on CCX282-B were in corticosteroid-free remission vs. 28% on placebo (p = 0.04). Compared to placebo, more subjects on CCX282-B had a normal CRP at 36 weeks (19% vs. 9%; p=0.04), and fewer subjects required corticosteroid rescue therapy (11% vs. 21%, p=0.04). CCX282-B was well tolerated. No serious or opportunistic infections occurred on CCX282-B. Serious adverse events occurred in 9% of subjects on CCX282-B and 10% on placebo. Conclusion: CCX282-B maintained remission in Crohn's disease over a 36-week period. This novel oral inhibitor of leukocyte trafficking may therefore represent an important new treatment option for the long-term management of inflammatory bowel dis- ease. 648 Randomised Placebo-Controlled Trial of Rituximab (Anti-CD20) in Active Ulcerative Colitis Keith Leiper, Kate Martin, Anthony Ellis, Sreedhar Subramanian, Alastair J. Watson, Stephen Christmas, Deborah Howarth, Fiona Campbell, Jonathan M. Rhodes Background Ulcerative colitis (UC) is associated with circulating auto-antibodies including pANCA. Although these are often seen as epiphenomena, they might be involved in pathogen- esis. Suppression of the humoral response by use of anti-B lymphocyte antibodies might S-86 AGA Abstracts then be therapeutic. Prompted by the efficacy of the B cell (anti-CD20) antibody, rituximab, in rheumatoid arthritis and by a good response in a patient with UC and lymphoma we have conducted a phase II double-blinded placebo-controlled study of this antibody in UC (NCT00261118). Method 24 patients with active UC (Mayo score 6-12, but excluding severe colitis as defined by Truelove and Witts) who had either failed to respond to 2 weeks systemic corticosteroids or who had relapsed during steroid tapering were randomised to receive rituximab 1g in 500mls 0.9% saline iv over 4h at 0 and 2 weeks (n=16) or saline placebo (n=8). Patients still receiving corticosteroids on entry (placebo group 7/8; rituximab group 14/16) continued a standard tapering regimen. Primary endpoint was remission (Mayo score < 2) at 4 weeks. Secondary endpoints included response (reduction in Mayo score by > 3 points) at 4 weeks and 12 weeks. The study was primarily designed to assess safety but was powered to give 80% chance of excluding an 80% remission rate with active treatment compared with an estimated 25% placebo rate. Patients were randomised in blocks of five and analysis was by Chi square, student's T or MannWhitney u, where appropriate. Results Mayo score on entry was significantly higher in rituximab-treated patients (mean 9.19; 95%CI 8.31-10.06) than for placebo patients (7.63; 6.63-8.62, p=0.03). Five of 8 placebo patients and 12 of 16 rituximab patients were receiving Azathioprine, 6-mercaptopurine or methotrexate (n=1). At week 4 only 1/8 placebo-treated patients and 3/16 rituximab patients were in remission. However at week 4 8/16 rituximab-treated patients had shown a response compared with 2/8 placebo with a median reduction in Mayo score of 2.5 (rituximab) compared with 0 (placebo; p=0.07). This response was however only maintained out to week 12 in 4/16. By week 12, UC had flared in 11/16 Rituximab patients and 7/8 placebo patients. Mucosal healing was seen at week 4 in 5/16 rituximab-treated patients and 2/8 placebo (NS). Rituximab was well tolerated with one chest infection, three mild infusion reactions plus one case of (probably unrelated) non-fatal pulmonary embolism. Conclusions Rituximab is well tolerated by patients with ulcerative colitis and may have some therapeutic effect. Larger studies, probably using a different dosing regimen, will be needed to assess this. 649 A Randomized Placebo-Controlled Trial of Abatacept for Moderately-to- Severely Active Crohn's Disease (CD) Stephen B. Hanauer, William J. Sandborn, Bruce E. Sands, Paul J. Rutgeerts, Remo Panaccione, Brian Bressler, Marvilyn A. Whiteside, René Swanink, Richard Aranda, Allison Luo Purpose: Abatacept selectively modulates the CD80/86:CD28 co-stimulatory signal required for T-cell activation. This placebo (PBO)-controlled trial evaluated the efficacy and safety of abatacept in patients with moderate-to-severely active CD and an inadequate response and/ or intolerance to medical therapy, including TNF antagonists. Methods: 451 adult patients with Crohn's Disease Activity Index (CDAI) ≥220 and ≤450, and hsCRP > upper limit of normal were randomized to abatacept 3 mg/kg, ~10 mg/kg, 30/~10 mg/kg or PBO; and received study drug during Induction Period (IP) on Days 1, 15, 29 and 57 (the 30/~10 mg/kg group received 30 mg/kg on Days 1 and 15; and ~10 mg/kg thereafter). Patients who met response criteria at Day 85 entered the Maintenance Period (MP; N=90) and were re randomized to ~10 mg/kg abatacept or PBO every 28 days. Primary IP endpoint was clinical response (reduction from baseline in CDAI of ≥100 points, or absolute CDAI <150 points) at both Days 57 and 85. Secondary IP endpoints included clinical remission (CDAI <150 points). The MP was prematurely terminated due to lack of efficacy in IP. Results: The proportions of patients with a clinical response at both Days 57 and 85 were 15.5, 10.2, 17.2 and 14.4% in the abatacept 3 mg/kg, ~10 mg/kg, and 30/~10 mg/kg, and PBO groups, respectively. The proportions of patients in clinical remission at both Days 57 and 85 were 4.7, 3.9, 7.8 and 6.3% in the abatacept 3 mg/kg, ~10 mg/kg, and 30/~10 mg/kg, and PBO groups, respectively. For the abatacept and PBO groups, clinical response rates were 35.1 and 40.9% at Month 6, 26.2 and 15.6% at Year 1, respectively. During IP, serious adverse events (SAEs) were reported in 15.4 to 17.2% of patients in the 3 abatacept treatment groups versus 15.6% of patients in the PBO group; serious infections (SIs) were reported in 3.1 to 7.0% of patients in the 3 abatacept treatment groups versus 2.3% of patients in the PBO group. During MP, SAEs were reported in 11.4% of abatacept and 19.6% PBO- treated patients; SIs were reported in similar proportions of abatacept and PBO-treated patients (2.3 and 2.2%, respectively). Other aspects of safety, including the proportion of patients with infections, were comparable across treatment groups in both study periods. Conclusion: Abatacept was not effective for induction of response and remission in patients with active CD. Noting the small number of patients treated and assessed in the MP due to the premature study termination, abatacept did not appear to demonstrate clinical activity in CD during the MP. The overall safety profile was similar between the treatment groups during both the IP and the MP. 650 Strain-Specific Downregulation of MicroRNA-320 by Carcinogenic Helicobacter pylori Promotes Expression of the Cell Survival Protein, MCL-1 Jennifer Noto, Courtney Bartel, Elizabeth J. Thatcher, James G. Patton, Richard M. Peek Colonization by Helicobacter pylori is necessary, but not sufficient, for progression to gastric cancer, emphasizing the importance of host cell responses to this pathogen. H. pylori strain 7.13 induces gastric cancer in rodent models of gastritis and was derived via In Vivo adaptation of a clinical strain, B128, which induces inflammation, but not cancer, thus providing a unique opportunity to define mechanisms that mediate carcinogenesis. MicroRNAs (miRNAs) are small noncoding RNAs that regulate gene expression post-transcriptionally via direct binding to the 3' untranslated region of messenger RNAs. miRNAs can function as either oncogenes or tumor suppressors and are frequently dysregulated in carcinogenesis. To identify miRNAs involved in H. pylori-mediated gastric cancer, miRNA microarray analyses were performed on total RNA isolated from MKN28 gastric epithelial cells co-cultured with H. pylori strain 7.13 or strain B128. Overall, 143 miRNAs were differentially expressed greater than 2-fold. The expression of a tumor suppressor miRNA, miR-320, was downregulated 17- fold in response to strain 7.13 and these results were confirmed by Northern blot analyses. Mcl-1, a Bcl-2 family member, has been reported to be negatively regulated by miR-320 and functions to inhibit apoptosis. To analyze Mcl-1 mRNA and protein levels in this model,