Long-term outcome of infliximab treatment in chronic active ulcerative colitis: a Swedish multicentre study of 250 patients.

BACKGROUND: Real-life long-term data on infliximab treatment in ulcerative colitis are limited. AIM: To study the long-term efficacy and safety of infliximab in chronic active ulcerative colitis and possible predictors of colectomy and response were also examined. METHODS: A retrospective multi-centre study of infliximab treatment in 250 patients with chronic active ulcerative colitis with inclusion criteria: age ≥18 years, ambulatory treated, steroid-dependent or intolerant and/or immunomodulator refractory or intolerant. RESULTS: Steroid-free clinical remission was achieved by 123/250 patients (49.2%) at 12 months and in 126/250 patients at a median follow-up of 2.9 years (50.4%). Primary response at 3 months was achieved by 190/250 (76.0%) patients and associated with a high probability of response 168/190 (88.4%) at 12 months and 143/190 (75.3%) at follow-up. Long-term rate of colectomy in primary responders was 6/190 (3.2%) at 12 months and 27/190 (14.2%) at last follow-up. Failure to achieve response at 3 months was associated with a high risk of subsequent colectomy, 29/60 (48.3%) at 12 months and 41/60 (68.3%) at follow-up. Response at 12 months was associated with a low risk of subsequent colectomy, 14/181 (7.7%) compared with non-response 19/34 (55.9%) (P < 0.0001). Non-response at 3 months was an independent predictor of subsequent colectomy (HR = 9.40, 95% CI = 5.10-17.35, P < 0.001). Concomitant azathioprine therapy did not influence outcome in terms of colectomy. CONCLUSIONS: Long-term efficacy of infliximab treatment in chronic active ulcerative colitis is excellent especially in patients who respond to induction treatment. Conversely, non-response at 3 months predicts a poor outcome, with a high risk of subsequent colectomy.

Large neural network simulations on multiple hardware platforms.

To efficiently simulate very large networks of interconnected neurons, particular consideration has to be given to the computer architecture being used. This article presents techniques for implementing simulators for large neural networks on a number of different computer architectures. The neuronal simulation task and the computer architectures of interest are first characterized, and the potential bottlenecks are highlighted. Then we describe the experience gained from adapting an existing simulator, SWIM, to two very different architectures-vector computers and multiprocessor workstations. This work lead to the implementation of a new simulation library, SPLIT, designed to allow efficient simulation of large networks on several architectures. Different computer architectures put different demands on the organization of both data structures and computations. Strict separation of such architecture considerations from the neuronal models and other simulation aspects makes it possible to construct both portable and extendible code.