Sa1881 Kinetics of the anti-HBs Titers After Vaccination Against Hepatitis B Virus (HBV) in Patients With Inflammatory Bowel Disease (IBD) Javier P. Gisbert, José R. Villagrasa, Amelia Rodriguez Nogueiras, Maria Chaparro Background: Among immunocompromised patients who respond to the vaccine, clinically significant HBV infection has been documented in those who do not maintain anti-HBs concentrations =10 IU/l. Objectives:To understand the kinetics of the anti-HBs titers over time in IBD patients who have responded initially to the vaccination. To identify predictive factors of negativization of anti-HBs titers. Methods: Patients from our IBD Unit with anti-HBs>10IU/l after HBV vaccination were prospectively included. Vaccination protocol consisted on a double dose of Engerix B® at 0, 1 and 2 months. Anti-HBs titers were measured 1-3 months after the last vaccine dose. Patients with anti-HBs <100IU/l after the 1 st attempt were considered failures and were vaccinated again with the same protocol.Patients with anti-HBs>10 UI/l after the 1 st or 2 nd vaccination were followed-up. Anti-HBs titers were measured at 6 and 12 months. When anti-HBs titers were lower than 10 UI/l during the follow-up, they were considered negatives. Long-term maintenance of positive anti-HBs titers was estimated using Kaplan-Meier curves. Cox-regression analysis was performed to identify potential predictive factors for losing anti-HBs protective titers. Results: 100 patients were included (mean age 42 years, 53% females, and 68% Crohn's disease).49% were on thiopurines, and 14% on anti-TNF drugs. 33 patients (33%) had received 2 vaccination attempts.Basal anti-HBs titers were>100 UI/l in 88% of patients. The cumulative incidence of negativization of the anti-HBs titers was 2% after 6 months and 15% after 12 months from the basal determination. The incidence rate of negativization of the anti-HBs titers was 18% per patient-years. Anti-HBs basal titers were statistically significant lower among patients who had negative values thereafter during the follow-up, compared with those who main- tained them over 10 IU/l (191 IU/l vs. 515 IU/l, p<0.001). In the multivariate analysis, neither the patient's age, gender, type of IBD or the treatment with thiopurines were associated with a higher probability of negativization of anti-HBs titers. The treatment with anti-TNF drugs was the only factor associated with a higher risk of negativization of anti-HBs titers (HR=3.1, 95%CI=1.1-8.8, p=0.03). Conclusions: A high proportion of IBD patients with protective anti-HBs titers after vaccination lose them over time (18% of patients per year of follow-up). The risk of losing protective anti-HBs titers is 3-fold higher among IBD patients on anti-TNF therapy. Sa1882 The New Immunoassay for the Accurate Determination of Antibodies to Infliximab, and Relationship Between Its Serum Level and Inflammatory Values in Crohn's Disease Hirotsugu Imaeda, Akira Andoh, Hiromitsu Ban, Shigeki Bamba, Masaya Sasaki, Tomoyuki Tsujikawa, Yoshihide Fujiyama [Background] The formation of antibodies to infliximab (ATIs) is closely associated with the loss of a response to infliximab in patients with Crohn's disease (CD). But the conventional method to measure ATI is including that many patients are diagnosed to false negative, because much of them are saturated by serum free infliximab. We evaluated the clinical utility of a novel method to measure serum ATI levels in the presence of infliximab. [Methods] 58 patients with CD under infliximab maintenance therapy were evaluated. The protocol of this study was approved by the Ethics Committee of the Shiga University of Medical Science. ATI levels were measured by a novel method of immunoassay with protein A and the conventional method. The serum infliximab trough levels were determined by enzyme- linked immunosorbent assay. [Results] ATIs were detected in 16 out of 58 patients (27.6%) by the new method, but the conventional method detected only 2 patients (3.4%) who had the 2 highest ATI titers assayed by the new method. The presence of ATIs in those samples positive according to the new method but negative by the conventional method was confirmed by Western blotting analysis. It also indicated that the new method could restore the binding capacities of the ATIs to infliximab, which were occupied by free infliximab. In the current method, the addition of infliximab to the samples dose-dependently blocked the detection of ATIs. Patients positive for ATIs had significantly lower serum trough levels of infliximab (p<0.01), and significantly higher clinical activity scores (p<0.001) and higher laboratory markers for inflammation (CRP and ESR) (p<0.01) as compared to patients negative for ATI. [Discussion] This new method to measure serum ATIs is useful more than conventional one. And the existence of ATI infliximab administrated CD cause loss of the effect of infliximab. So we established the strategy for second failure to infliximab. Patients whose serum Infliximab levels are high, it is not thought to be caused by TNF-alpha, they should be administrated of another medication like immunomodulator or prebiotics. Some of patients whose serum Infliximab levels are low and whose ATI levels are high, must be changed to Adalimumab. The others whose ATI levels are low should be increased the dosage of infliximab. [Conclusion] The current method makes it possible to measure serum ATI levels in the presence of infliximab. This new method seems to be useful for deciding the optimal management strategies for CD patients with a loss of response to infliximab. Sa1883 Long-Term Evolution of Crohn's Disease Patients Responders to Azathioprine Marine Camus, Philippe Seksik, Anne Bourrier, Isabelle Nion-Larmurier, Harry Sokol, Philippe Baumer, Laurent Beaugerie, Jacques Cosnes Steroid-free clinical remission at one year is obtained in 30-50% of patients with Crohn's disease (CD) treated with Azathioprine (AZA). Long term outcome of patients who achieved such a response (AZA responders) is not well known. Some patients may lose response and safety during prolonged treatment with AZA. The aim of our study was to assess the long term efficacy and tolerance of AZA in AZA responders Methods: we selected retrospectively from the MICISTA registry 220 patients who had been placed under AZA between 1987 and 1999 and achieved a complete response to AZA (steroid-free clinical remission at 1 year of treatment: 41% of patients who received AZA during the same period). Clinical S-349 AGA Abstracts relapses, year-by-year disease activity, hospitalizations, need for intestinal and perianal sur- gery, need for anti-TNF treatment, and occurrence of serious adverse events had been collected prospectively in the registry during follow-up. Evolution was compared to that of CD patients from the registry who, during the 1987-99 period, were not placed under immunosuppressants. Each AZA responder was matched to two CD controls by gender, year of birth, and year of diagnosis of CD. Results: AZA responders (86 M, 134 F, median age [IQR] 32 years [25-39] were followed-up for a median (IQR) of 12.6 years (10.3-15.4), i.e. 14 years after starting AZA. AZA was never stopped in 106 patients, stopped till last follow-up in 73 patients, and stopped for 0.5-8 years then resumed in 41 patients. The probabilities of sustained clinical remission and remaining free of any surgery or antiTNFα after 10 years of follow-up were 0.38 and 0.64, respectively. Independent predictors of clinical relapse were younger age (OR 95%CI 1.45 [1.39-1.53]) and prior perforating com- plication (OR 95CI% 1.67 [1.58-1.75]). Compared with the control group (187 patients [42%] eventually received immunosuppressants after inclusion), AZA responders had a lesser year-by-year disease activity (20 vs. 29% p<0.001), less years with hospitalization (6 vs. 9%, p<0.001), less years with intestinal surgery (2.8 vs 4% p<0.01) and less years with perianal surgery (8.6 vs. 19.5% p <0.001). Requirement of antiTNF was not different (8.7 vs 7.5%). Eleven AZA responders (5%) eventually died, vs. 7 controls (1.6%, p<0.05). Twenty-three AZA responders (9.5%) developed cancer (7 intestinal, 2 lymphoma, 5 non melanoma skin cancer, 10 others), vs. 23 controls (5.2%, p=0.04) (8 intestinal, 3 lymphoma, 1 non melanoma skin cancer, 10 others). Eight AZA responders (3.6%) died from cancer, vs. five controls (1.1%, p=0.03). Conclusion: In responders, AZA does maintain its efficacy during the long-term, decreasing, albeit not to zero, disease activity and the need for surgery (both intestinal and perianal). However AZA responders are exposed to an increased risk of malignancy and a prolonged careful monitoring is mandatory. Sa1884 Infliximab Versus Immunomodulator as First Maintenance Therapy in Children With Crohn Disease James Markowitz, Jonathan Evans, Marian D. Pfefferkorn, Anne M. Griffiths, David J. Keljo, David R. Mack, Ryan Carvalho, Neal S. Leleiko, Anthony R. Otley, Joel R. Rosh, Michael C. Stephens, Shehzad A. Saeed, Farhat N. Ashai-Khan, Athos Bousvaros, Marsha H. Kay, Maria Oliva-Hemker, Andrew B. Grossman, Boris Sudel, Michael Kappelman, Reed A. Dimmitt, Christine R. Langton, Jeffrey S. Hyams Background: Despite widespread speculation regarding potential benefits of top-down inflixi- mab (ifx) therapy for Crohn disease (CD), little pediatric data have been reported. Methods: A retrospective cohort study from the prospective database of the Pediatric IBD Collaborative Research Group Registry was undertaken. Children are enrolled in the Registry at diagnosis (dx), and are treated by physician's judgement not protocol. Data are collected at dx, 30 days after dx and quarterly thereafter. Subjects for this analysis included all children receiving ifx within the first 30 days after dx of CD and followed for ≥1 yr. Each subject was matched to 2 children not treated with ifx in the 1st 30 days, using criteria that included age at dx (±1 yr), year of dx (±3 yr), CD activity at dx (mild, moderate-severe) and extent of CD (L1, L2, L3). Duration of a triad's followup (f/u) was restricted to the shortest duration noted for any member of the triad. Results: 37 ifx subjects (Grp1) were matched to 71 controls (Grp2) (3 subjects could only be matched to 1 control). Grp2 subjects received a thiopurine (n=62) or methotrexate (n=9). Age at dx (12.5 ± 2.2 yr) and mean duration of f/u (30 months) were identical, and both groups had similar CD activity at dx (moderate-severe [Grp1: 84%, Grp2: 93%]). Baseline WBC, Hgb, and ESR were comparable between groups, but albumin was lower in Grp 1 (2.9 ± 0.7 vs 3.3 ± 0.7, p=0.0322). Corticosteroid (CS) use in the 1st 30 days was more common in Grp2 (70% Grp1, 89% Grp2, p=0.03), but months on CS over the f/u period were similar (6.2 ± 5.6 Grp1, 6.9 ± 5.3 Grp2, p=NS). At 1 yr, rates of CS free, inactive or mild CD were similar between groups (79% Grp1, 81% Grp2). Subjects with at least 1 hospitalization were significantly more common in Grp1 (54% Grp1, 25% Grp2, p=0.0053), but this was predominantly due to subjects with an initial hospitalization in the 1st 30 days (30% Grp1, 7% Grp2, p=0.0032). Subjects with 2 or more hospitalizations were similar between groups (40% Grp1, 44% Grp2). Both groups had similar rates of subjects with at least 1 intestinal surgery (8% Grp1, 14% Grp2) and at least 1 perianal surgery (5% Grp1, 7% Grp2). Among Grp1 subjects, 22/37 (51%) remained on ifx throughout f/u, while 22% switched to a 2nd biologic and 19% to an immunomodul- ator. In Grp2, 36/71 (51%) remained on their initial immunomodulator, 11% were switched to a 2nd immunomodulator and 38% to ifx. Conclusion: There was little difference in measured outcomes between children receiving either ifx or an immunomodulator as first maintenance therapy over a mean of 30 months observation. However, lower mean serum albumin levels and higher hospitalization rate in the 1st 30 days suggest a treatment bias toward selecting somewhat sicker subjects for ifx. Sa1885 Infliximab Trough Levels in Patients' Developping Cutaneous and Rheumatologic Paradoxical Manifestations: A Case-Control Study David Laharie, Julie Chapuis, Edouard Chabrun, Valerie Labat-Debelleix, Maylis Capdepont, Patrick Blanco, Victor de Ledinghen Introduction: Patients with an inflammatory bowel disease (IBD) treated with infliximab (IFX) may develop cutaneous or rheumatologic paradoxical manifestations (CPM and RPM, respectively) with time. Their pathogenesis remains unclear and may be associated with high IFX trough levels (ITL). The aim of our case-control study was to identify an association between ITL values and the occurrence of CPM or RPM. Patients & methods: From May 2010 to January 2011, IBD patients receiving maintenance with IFX were consecutively included in a monocenter cross-sectional study. At inclusion, defined by the first IFX infusion as maintenance during the study period, patients were screened for CPM and RPM. Diagnosis of paradoxical manifestation was assessed by the occurrence of typical clinical lesions and/ or symptoms in patients without any cutaneous lesions or rheumatism before starting IFX, secondary confirmed by a dermatologist or rheumatologist. ITL, anti-IFX antibodies (ATI) (LISA-TRAKER®, Biomedical Diagnostics BMD) and antinuclear antibodies (ANA) were measured at baseline. ITL values were compared between patients with CPM or RPM and AGA Abstracts