primary end-point was the annual remission rate. Patients were considered in remission if CDAI was (150 . Secondary end-points were a) time in the study, b) time to relapse and c) QoL ~core. Results: One hundred patients were enrolled, 58 in group A and 42 in group B. There were no baseline differences in any patient or disease-related parameters between study groups. The mean duration of treatment was 3.3 (range 2.2-3.9) years in group A and 6.9 (range 5.3-8) in group B. By the end of the study 8/58 (13.8%) of patients in group /~ and 5/42 (12%) in group B had been withdrawn (P = 1.0). Two patients from group A and one from group B were withdrawn because of drug-related side effects. Thus, the PP relapse rate was 10.7% and 9.7%, respectively (P = 0.8). The mean daily dose of azathioprine during the study was 145mg for group A and 143.5mg for group B. There were no significant differences in relapse rates between patients with the same disease location but different duration of treatment. Multifactorial analysis did not reveal any patient or disease related parameter that influenced the one year outcome of the study between groups A and B. There were no significant differences overall and at each time point in both CDAI and IBD- Q scores between study groups. At the end of the study the Crohn's Disease Endoscopic Index of Severity (CDEIS) was 3.2(2.6) and 3.57(2.1) (mean, range) for groups A and B, respectively (P = 0.5). Conclusion: Providing that patients are comply with treatment, the efficacy of azathioprine in maintaining remission of CD does not appear to wane off with duration of treatment.

M1584

Leukocytapheresis in Ulcerative Colitis - First Results of a Controlled Study Joerg Emmrich, Dietrich Nowak, Peter Brock, Gerd Beumer, Sebastian Petermann, Wolfgang Ramlow, Rainhard Klingel, Stefan Liebe

Introduction: Recently published studies have suggested that leukocytapheresis (LCAP) is a useful therapy of inflammatory bowel disease after failure of conventional treatments. Methods: We recruited 21 patients with active ulcerative colitis (CAI according Rachmilewitz: 6 - 10) treated with steroids (at least 10 mg/die prednisolon) for more than 6 months. Two weeks before and during the study steroid dosage was constant 10 mg prednisolon per day. LCAP was performed weekly for five weeks (Intensive Therapy) using column Cellsorba FX (Asahi Medical, Tokyo, Japan). After this treatment patients with CAI ~ 4 were randomized in two groups followed by tapering the steroids. In one group LCAP was continued monthly for 5 months (Maintenance Therapy) whereas patients in the second group were not treated with LCAP. Lymphocyte subpopulations were monitored before and after LCAP using flow cytometry. Results: Remission of disease (CAI ~ 4) was achieved in 14 of 21 patients after Intensive Therapy with LCAP. Up to now 11 of these 14 patients have terminated the study. After the Maintenance Therapy 3 of 6 patients had no disease activity. 2 of 5 patients treated with Intensive Therapy without following Maintenance Therapy were still in remission after 5 months. LCAP reduced the number of T- lymphocytes (CD3+, 31.1%), activated T lymphocytes (CD3+, HLA DR+, 26.4%), and also B lymphocytes (CD19, 28,2%) in the peripheral blood. Conclusion: LCAP was successful to induce remission in patients with ulcerative colitis. There were no LCAP - related side effects. T cells as well as B lymphocytes were removed by the used apheresis technique. Further studies are needed to evaluate the optimal treatment protocol and the influence of LCAP on the immune system in the gut.

M1585

Quali ty of Life Improvements in a Phase 2 Study of Natal izumab for Active Crohn's Disease Paul Rtugeerts, Stephen Donoghue, Tanya Palmer

Introduction: The Inflammatory Bowel Disease Questionnaire (IBDQ) was used to assess heafih-related quality of life changes in a randormzed, double-blind, placebo-controlled study of natalizumab in 248 patients with moderate to severely active Crohn's disease. Natalizumab, a product in the class of selective adhesion molecule inhibitors (SAM inhibitora), is a humanized monoclonal antibody to ct4 integrin, previously reported to have an effect on both disease remission (Crohn's Disease Activity Index (CDAI) ~ 150) and clinical response (at least a 70 point reduction in CDA1 from baseline) in this trial. Methods: Responses to the IBDQ were examined for all patients (intention to treat). The 4 treatment groups received either a single 3 mg/kg natalizumab intravenous infusion; two 3 mg/kg natalizumab infusions at a 4-week imerval; two 6 mg/kg natalizumab infusions at a 4-week interval; or placebo. The overall IBDQ score and individual dimension data were analyzed using the Wilcoxon-Mann-Whitney test to compare the change from baseline between each of the active treatment groups and placebo at week 6. Results: Patients treated with natalizu- mab (any dose) had statistically significant improvements in the median 1BDQ score at week 6 when compared to patients treated with placebo (3mg/kg group, P=0.008, 3 + 3 mg/kg group, P~0.001, 6 + 6 mg/kg, P~0.001). Patients receiving 2 infusions of natalizumab also had statistically greater improvements in all 4 (bowel, social, emotional, and systemic) IBDQ dimensions compared to placebo (P~0.05 for all groups receiving 2 doses of natalizumab for each dimension). Conclusions: Quality of life of patients with moderate to severely active Crohn's disease receiving natalizumab was improved as assessed by patient responses to the IBDQ in this short-term study. These results suggest that natalizumab may hold promise as a treatment for Crohn's disease. Ongoing studies are currently underway to confirm and extend these findings.

M1586

Relationship of Antibodies to Infliximab to Durat ion of Response and Infusion Reactions for Crohn's Disease Patients Receiving Episodic Infliximah Retreatment in ACCENT I Carrie Wagner, Stephen B. Hanauer, Mohan Bala, Kamlesh Patel, Suzanne Travers, Allan Olson, Paul Rutgeerts

Background: Efficacy and safety of infliximab (IPX) retreatment were compared for patients (pts) positive (pos), negative (neg) and inconclusive (incoucl) for antibodies (Ab) to 1FX at week (wk) 14 following a single 5 mg/kg 1FX treatment at week 0. Methods: All pts in

ACCENT I received 5 mg/kg IFX at week 0. Patients randomized to the placebo group received placebo at wks 2 and 6 then every 8 weeks. Starting at wk 14, pts losing response in placebo group were eligible to receive episodic retreatment (retx) with 5mg/kg 1PX. Ab to IFX were measured using a double anugen EtA, and a monoclonal EtA was used to measure serum IFX levels. Pts without Ab to IFX detected were classified as either neg (no serum IFX) or inconcl (detectable IFX). Duration of response was based on observed Crohn's Disease Activity Index (CDAI) at each visit relative to baseline. Patients with medication changes or surgeries were considered as having lost response. Results: At wk 14, 24 (14%) pts were Ab pos, 100 (56%) were Ab neg, and 53 (30%) were inconcl. Similar proportions of Ab pos (58%) and Ab neg (59%) pts crossed over to episodic retx at week 14 or later, compared with 32% of inconcl pts. Similar proportions ofpts responded to the first crossover infusion (86%, 85% and 94% of pos, neg and inconcl pts, respectively), and similar propor- tions remained in clinical response at week 54 (63%, 75% and 70%, respectively). The average duration of response was 13.5, 17.5 and 31 wk, for Ab pos, neg and inconcl pts, respectively. The average number of crossover infusions (1.50 versus 1.25) over the 1 year period, and the average duration of study participation (44 wk versus 39 wk), was similar between Ab pos and neg groups. Inconcl pts had fewer (0.74) crossover infusions, but had a similar duration of study involvement (46 wk). A higher incidence of generally nonserions infusion reactions occurred at the first crossover infusion for Ab pos pts (28.6%) relative to the neg (5%) and inconcl pts (6%). Conclusions: A higher incidence of mild to moderate infusion reactions was observed for pts who were pos for antibodies to IFX compared with neg or inconcl pts. Despite a shorter duration of response for the Ab pos pts, most pts remained in the study, responded to retx and were in clinical response at the end of the study.

M1587

Budesonide for the Induction of Remission in Crohn's Disease: Meta-Analysis of Randomized Controlled Trials Anthony Otley, Alan B. Thomson, Robert Modigiliani, Ole O. Thomsen, Hiflary Steinhart

Introduction: Corticosteroids are the most effective first-line treatment for acute flares of Crohn's disease but they can induce sigmficant side effects. Budesonide is a corticosteroid with a high anti-inflammatory potency, but low systemic bioavailabifity when given orally or rectally because of extensive degradation (90%) at first pass hepatic metabolism. Objective: To conduct a meta-analysis to determine the efficacy and safety of budesonide for the induction of remission in active Crohn's disease. Methods: A MEDLINE search from 1986 onwards, and hand-searching of references from articles and abstracts from major gastrointes- tinal society meetings was performed. Selection criteria included patients of any age with acutely active Crohn's disease, as defined by a CDAI > 150; randomized, double blind controlled trials comparing treatment with budesonide to a control treatment (patiems in the control arm may have received placebo, conventional corticosteroids, 5-aminosalicylic acid or suffasalazine); clinical remission was the outcome of interest as defined by a CDAI < 150 by 8 to 16 weeks of therapy, secondary outcomes included corticosteroid-related side effects (CRSE). Data was extracted by two independent reviewers and the quality of each paper was also assessed. Analysis was performed using the Revman statistical software to calculate pooled odds ratio (OR) and chi-square values in the test nf heterogeneity Results: Eight studies were identified which met selection criteria. Budesonide was less likely to induce a remission of active Crohn's disease than conventional corticosteroids (OR 0.69, 95% CI: 0.51-0.95). However, compared with both mesalamine (OR 2.80, 95% CI:1.50- 5.20) or placebo (OR 2.85, 95% CI: 1.67-4.87) budesonide was more likely to induce a remission of active disease. The proportion of patients reported to have at least one CRSE was sigmficantly less in the budesonide group compared with those receiving conventional steroids (OR 0.38, 95% CI:0.28-0.53). No significant difference in the proportion of CRSE was noted between budesonide and placebo-treated groups (OR 0.98, 95% C1:0.58-1.67) at 8 weeks. Conclusions: When compared with conventional corticosteroid, budesonide is less efficacious for induction of remission of active Crohn's disease. However, budesonide is more efficacious when compared to mesalamine or placebo and over the usual acute treatment period of eight weeks was not associated with corticosteroid-related side effects.

M1588

Regardless of Duration of Corticosteroid Treatment, lnfliximab (REMICADE) Maintenance Therapy Enables Steroid Elimination: ACCENT I Experience Gary R. Lichtenstein, Allan Olson, Robert Diamond, Warren Bao, Stephen B. Hanauer

Background: lnffiximab has been shown to induce and maintain clinical remission and enable reduction or elimination of corticosteroids in patients with Crohn's disease (CD). The objective of this presemation was to examine whether duration of steroid use affected the ability of infliximab to reduce or eliminate steroids, using ACCENT I data. Methods: A total of 573 patients with a CDAI score between 220 and 400 received a single infusion of 5 mg/kg infliximab. At week 2, patients were randomized to 1 of 3 maintenance regimens: (Group I) placebo at weeks 2 and 6 and then placebo every (q) 8 weeks; (Group II) 5 rag/ kg infiiximab at weeks 2 and 6 and then 5 mg/kg q8 weeks; and (Group Ill) 5 mg/kg nifliximab at weeks 2 and 6 and then 10 mg/kg infliximab q8 weeks. After week 6, patients who showed clinical response could have corticosteroids tapered at a rate of 5 rag/week if the baseline dose was > 20 rag/day prednisone and 2.5 rag/week if the baseline dose was -< 20 mg/day prednisone. In patients who took corticosteroids at baseline and were responders at week 2, the steroid sparing benefit of infiiximab was assessed by duration of steroid use prior to study entry (< 1 year or > 1 year). Results: Approximately 52.2% of the week 2 responders were receiving corticosteroids at baseline. Dose at baseline between treatmem groups, in both patients who used steroids for -< 1 year and for >1 year. lnfliximab maintenance treatment enabled early reduction and elimination of steroids. OveraU, 8.9%, 24.1%, and 32.1% of patients in Groups I, If, and Ill, respectively, were able to taper off steroids while remaining in remission. Conclusions: Regardless of duration of steroid use prior to infliximab therapy, induction and maintenance treatment with infliximab enables steroid reduction and elimination. In patients with -< 1 year of steroid use at baseline, maintenance therapy was 3 times more likely to eliminate steroid use. In all treatment groups, patients with -< 1 year of corticosteroid use were more likely to reduce and elimi- nate corticosteroids.

AGA Abstracts A-378