P461 Endoscopic features for loss of response in patients with Crohn’s disease who were treated with infliximab by top-down strategy

The top-down strategy of treatment with anti-TNF agents showed the potentials for improved efficacy and outcomes in patients with Crohn’s disease (CD), especially those suspected to have a poor prognosis. However, few studies have evaluated clinical and endoscopic features associated with secondary loss of response (LOR) in CD cases treated with infliximab by the top-down strategy. We treated 410 CD patients with infliximab (IFX) from December 2004 to May 2010 in our hospital. Among these CD cases treated with IFX, those receiving the top-down regimen were defined by a disease duration of less than 2 years, no treatment history of steroid/immunomodulator/biologics, and no history of surgery. Effectiveness of IFX was defined on the basis of a more than 70-point decrease in the Crohn’s disease activity index (CDAI). Endoscopic effectiveness was defined as a more than 50% decrease in the simple endoscopic score for Crohn’s disease (SES-CD). LOR was defined as a more than 50-point increase, requiring additional or increasing doses of concomitant therapy. We retrospectively investigated 58 CD cases treated with infliximab by the top-down strategy. The cumulative remission rate was 86.1% at 1 year, 70.0 at 2 years and 61.0% at 4 years. The LOR group (n = 24, 10.9 ± 8.8 months) had a significantly longer disease duration than the non-LOR group (n = 34, 9.9 ± 18.8 months) (p = 0.04). The other factors at baseline including concomitant immunomodulator administration, albumin level, CDAI, whole SES-CD and segmental SES-CD of 5 sections each did not differ between the 2 groups. Among the changes in segmental SES-CD at Week 52 from week 0 as the index of endoscopic improvement, the scores for the caecum and ascending colon (−2.0 ± 2.0), the descending and sigmoid colon (−1.0 ± 1.9) and the rectum (+0.4 ± 2.4) in the LOR group were significantly lower than those in the non-LOR group (−4.3 ± 0.5, p < 0.01; −5.3 ± 2.4, p < 0.01; −2.6 ± 1.8, p < 0.02), while endoscopic improvements of the other segments were observed in both groups. Especially, longitudinal ulceration in the descending and sigmoid colon in the LOR group was not significantly improved as compared with that in the non-LOR group (positive rate: 62.5% vs. 71.4% at week 0, 62.5% vs. 14.3% at Week 52; p = 0.05). LOR occurrence rate in CD patients treated with IFX by the top-down strategy was similar to that in CD patients treated with IFX by the conventional strategy. The existence of an active lesion in the ileocaecum or distal colon, especially a highly active lesion (eg, longitudinal ulcer) in the distal colon, at baseline might predict LOR endoscopically.