Su1120

Appropriateness and Long-Term Discontinuation Rate of Biological Therapiesin Ulcerative ColitisMichel H. Maillard, John-Paul Vader, Christian Mottet, Alain Schoepfer, Jean-JacquesGonvers, Bernard Burnand, Florian Froehlich, Pierre F. Michetti, Valérie Pittet

Background: Anti-TNFalpha agents are commonly used for ulcerative colitis (UC) therapyin the event of non-response to conventional strategies or as colon-salvaging therapy. Theobjectives were to assess the appropriateness of biological therapies for UC patients and tostudy treatment discontinuation over time, according to appropriateness of treatment, as ameasure of outcome. Methods: We selected adult ulcerative colitis patients from the SwissIBD cohort who had been treated with anti-TNFalpha agents. Appropriateness of the first-line anti-TNFalpha treatment was assessed using detailed criteria developed during theEuropean Panel on the Appropriateness of Therapy for UC. Treatment discontinuation asan outcome was assessed for categories of appropriateness. Results: Appropriateness of thefirst-line biological treatment was determined in 186 UC patients. For 64% of them, thistreatment was considered appropriate. During follow-up, 37% of all patients discontinuedbiological treatment, 17% specifically because of failure. Time-to-failure of treatment wassignificantly different among patients on an appropriate biological treatment compared tothose for whom the treatment was considered not appropriate (p=0.0007). Discontinuationrate after 2 years was 26% compared to 54% between those two groups. Patients oninappropriate biological treatment were more likely to have severe disease, concomitantsteroids and/or immunomodulators. They were also consistently more likely to suffer afailure of efficacy and to stop therapy during follow-up. Conclusion: Appropriateness offirst-line anti-TNFalpha therapy results in a greater likelihood of continuing with the therapy.In situations where biological treatment is uncertain or inappropriate, physicians shouldconsider other options instead of prescribing anti-TNFalpha agents.

Su1121

Anti-TNF and Anti-Drug Antibodies Levels Predict the Outcomes ofInterventions After Loss of Response to Adalimumab and InfliximabHenit Yanai, Lev Lichtenstein, Amit Assa, Yoav Mazor, Batia Weiss, Arie Levine, YuliaRon, Uri Kopylov, Yoram Bujanover, Yoram Rosenbach, Bella Ungar, Abraham R. Eliakim,Yehuda Chowers, Raanan Shamir, Gerald Fraser, Iris Dotan, Shomron Ben-Horin

Background & aims: The usefulness of measuring anti-TNF drug and anti-drug-antibodies(ADA) levels remains controversial. We investigated the correlation between drug/ADA levelsand the outcome of therapeutic interventions in patients with loss of response (LOR) toinfliximab or adalimumab. Methods: A multi-center retrospective study of pediatric andadult IBD patients with LOR to anti-TNFs. The outcomes of therapeutic interventions wereassessed in relation to drug/ADA trough levels and evidence of inflammation at the time ofLOR. Results: There were 247 IBD patients (42 ulcerative colitis) who experienced a totalof 330 LOR events (188 to infliximab, 142 to adalimumab). Unequivocal immunogenicity(drug-/ADA+) was more common in LOR to infliximab than to adalimumab (47% vs. 23%,respectively, OR 4.6,95% CI 2.1-10,p<0.001). Slightly more infliximab LOR events weremanaged by dose-intensification (48% vs. 37% for adalimumab, p=0.045), whereas no-intervention was a more common strategy in adalimumab-treated patients (28% vs. 7% forinfliximab, p<0.001). Adalimumab level>4.5mcg/ml and infliximab>3.8 mcg/ml were 90%specific for failure to respond to dose-intensification or to anti-TNF switch, and had >75%specificity for a favorable response to either no-intervention or to out-of-class interventions.Antibodies-to-adalimumab>4mcg/ml-eq and antibodies-to-infliximab>9mcg/ml-eq were90% specific for failure to respond to dose-intensification. Patients with high-titer ADA hadlonger response duration to a switch of anti-TNF compared to dose-intensification (p=0.03,log rank test), but dose-intensification was superior for patients with no/low-titer ADA (p=0.02) Conclusions: Predominant mechanisms of clinical LOR are somewhat different betweenadalimumab and infliximab. Drug/ADA levels may guide therapeutic decisions in over two-thirds of patients with LOR, possibly allowing expectant management in some patients.

Su1122

Monitoring C-Reactive Protein in Biologic Naive Patients With UlcerativeColitis Undergoing Infliximab Remission Induction TherapyYasuo Suzuki, Ryota Iwasa, Akihiro Yamada

BACKGROUND: In recent years, the anti-tumour necrosis factor (TNF)-α, infliximab (IFX)has been applied in the treatment of patients with moderate to severe ulcerative colitis(UC). However, a sub-group of UC patients do not respond to IFX. This investigation wasundertaken to identify background factors, which could mark a patient responder or other-wise a non-responder to IFX. METHODS: Seventy-two patients with the partial Mayo score(pMayo) of 4-9 and IFX naive received IFX induction therapy. The pMayo, trough IFX leveland C-reactive protein (CRP) were measured. IFX infusion (5mg/kg bodyweight) was doneat weeks 0, 2, 6, and clinical response was assessed at week 14 based on the pMayo. Changesin CRP were analyzed by using receiver operating characteristic (ROC) graphs. RESULTS:Clinical response was achieved in 32 of 72 patients (44.4%). Relative to baseline, CRP(mg/dL) level at week 2 in responders (median, interquartile range 0.07, 0.02-0.19) wassignificantly lower than in non-responders (0.21, 0.08-0.44, P=0.0370) or patients withworsened UC (0.65, 0.21-2.18, P=0.0022) in spite of similar trough IFX levels. The medianCRP (week 2/week 0) ratio was significantly lower in patients who responded to IFX vsnon-responders and those with worsened UC, 0.07 vs 0.39 and 1.19, respectively. A cut-off CRP ratio of 0.19 (week 2/week 0 ratio) on the ROC curve could predict non-responderswith an 82.5% sensitivity and 75% specificity. Further, the area under the ROC curve wasestimated to be 0.79 with a 95% confidence interval of 0.67-0.91 (P<0.001). CONCLUSIONS:In this study, a significant difference in CRP level between responders (lower CRP level)and non-responders or patients with worsened UC (higher CRP level) during IFX inductiontherapy was found 2 weeks after the first IFX infusion. Accordingly, we believe that evenin patients with similar trough IFX level, the CRP ratio is potentially interesting in clinical

S-381 AGA Abstracts

setting as determination of CRP at an early stage might indicate stopping futile anti-TNF-α therapy and spare medical resources.

Su1123

Top-Down Versus Step-Up Treatment in Newly Diagnosed Crohn's Disease: NoDifference in Long-Term OutcomeJudith A. Stibbe, Daniël R. Hoekman, Filip J. Baert, Philip Caenepeel, Philippe L.Vergauwe, Martine De Vos, Severine Vermeire, Geert R. D'Haens

Background: Early combined immunosuppression (‘top-down' (TD)) is more effective thanconventional management (‘step-up' (SU)) for induction of remission and reduction ofcorticosteroid use in patients recently diagnosed with Crohn's disease (CD). However, itremains unknown whether short-term benefits are sustained long-term and thus the naturalhistory of CD can be altered. Therefore, we aimed to investigate the long-term effects ofTD (induction IFX and maintenance azathioprine (AZA)) vs. conventional SU treatment inCD. Methods: Long-term follow-up data was retrospectively collected from patients whoparticipated in a randomized controlled trial evaluating TD vs. SU in patients with newlydiagnosed Crohn's disease (1). Data collection was performed in 12 of the 18 participatingcenters. For 16 semesters following the original trial follow-up, the following was abstractedfrom patients' medical records: clinical disease activity by global assessment, medicationuse, hospitalization, surgery, and the occurrence of new fistulas and significant flares.Comparisons were done by intention-to-treat analysis. Time to event data was evaluatedusing the Kaplan-Meier and log-rank test. To compare the proportions of time in remission,Fisher's exact test was used. Algorithm failure was considered any of the following: surgery,start of adalimumab, ciclosporin or experimental therapy. Results: 112 patients (SU n=57)were included in the analysis, of whom 83.5% was in clinical remission during ≥50% ofsemesters. At the start of follow-up, 81.6% (57.1% AZA, 24.5% methotrexate (MTX)) vs.66.7% (54.2% AZA, 12.5% MTX) of patients used an immunomodulator, and 20.4% vs.16.7% received IFX in TD and SU, respectively. No difference in the proportion of semestersthat patients were in clinical remission during follow-up was found between TD and SU(66.6% vs. 68.3%; p=0.52). Mean time to first hospitalization was 13.9 vs. 13.1 semesters(p=0.46), mean time to first new fistula was 15.1 vs. 14.5 semesters (p=0.53) and meantime to algorithm failure was 11.7 vs. 10.5 semesters (p=0.27). The median time to Crohn-related surgery was similar in both groups (15.0 vs. 14.1; p=0.28). A trend towards adifference between TD and SU was observed for time to significant flare (median time 6 vs.8 semesters; p=0.09). Conclusion: No difference in long-term outcome was found betweentop-down versus step-up treatment algorithms for newly diagnosed Crohn's disease. Apotential explanation may be that top-down induction was restricted to only 3 IFX infusions.Furthermore, an early start of immunomodulation and/or IFX in the SU group, could havelead to a reduced contrast between patient groups. (1) D'Haens, GR et al. Early combinedimmunosuppression or conventional management in patients with newly diagnosed Crohn'sdisease: an open randomised trial. Lancet 2008;371:660-667.

Su1125

Long-Term Efficacy of a pH-Dependent Release Mesalamine Formulation,Asacol in Patients With Ulcerative Colitis Who Showed Inadequate Responseto a Time-Dependent Release Mesalamine Formulation, Pentasa: A ProspectiveStudyNaoki Yoshimura, Takaaki Kawaguchi, Minako Sako, Abbi Saniabadi, Masakazu Takazoe

BACKGROUND: Hitherto, mesalamine has often been used as a first-line medication forinducing and maintaining remission in patients with mildly to moderately active ulcerativecolitis (UC). However, there is no evidence for efficacy difference between the pH-dependentrelease mesalamine formulation, Asacol and the time-dependent release formulation, Pentasa.We were interested to evaluate the efficacy of Asacol instead of low dose (≤2.25g/day),moderate dose (3.0g/day) or high dose (4.0g/day) Pentasa for inducing and maintainingremission in UC patients. METHODS: In a single-centre prospective setting, 180 consecutivepatients with mildly to moderately active UC who had shown inadequate response to Pentasaover a period of ≥4 weeks willingly switched from oral Pentasa (mean dose 2.9±0.8g/day;range 1.5-4.0/day) to Asacol at 3.6g/day. During this study, no patient received corticosteroidor immunosuppressant. At entry and week 4, patients were clinically and endoscopicallyevaluated. Clinical response was defined as Rachmilewitz's clinical activity index (CAI) ≤4,essentially CAI ≤1 meant clinical remission. Similarly, endoscopic remission was defined asRachmilewitz's endoscopic index (EI) ≤3, essentially EI≤1 meant mucosal healing. Patientswho responded to Asacol remission induction therapy received maintenance Asacol at 2.4gor 3.6g/day, and were followed for 52 weeks. The sustainability of response in patients whoresponded at week 4 was followed by Kaplan-Meier survival analysis. RESULTS: Of the 180

AG

AA

bst

ract

s