Predicting outcomes in acute severe ulcerative colitis

Response to corticosteroid treatment in acute severe ulcerative colitis (ASUC) has changed very little in the past 50 years. Predicting those at risk at an early stage helps stratify patients into those who may require second line therapy or early surgical treatment. Traditionally, risk scores have used a combination of clinical, radiological and biochemical parameters; established indices include the 'Travis' and 'Ho' scores. Recently, inflammatory bowel disease genetic risk alleles have been built into models to predict outcome in ASUC. Given the multifactorial nature of inflammatory bowel disease pathogenesis, in the future, composite scores integrating clinical, biochemical, serological, genetic and other '-omic' data will be increasingly investigated. Although these new genetic prediction models are promising, they have yet to supplant traditional scores, which remain the best practice. In this modern era of rescue therapies in ASUC, robust scoring systems to predict failure of ciclosporine and infliximab must be devised.