Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P164. Predictive factors of the course of Crohn's disease - can we treat in anticipation?

F. Dias de Castro, J. Magalhães, P. Boal Carvalho, M.J. Moreira, J. Cotter, Centro Hospitalar do Alto Ave, Gastroenterology, Guimaraes, Portugal


The course of Crohn's disease (CD) is highly variable and difficult to predict on the basis of information at the time of diagnosis. However it would be useful to categorize patients at the onset of disease in low and high risk on the basis that treating severe CD with “top-down” strategy might change the natural history of CD.

The aim of this study was to evaluate factors at presentation that might predict the severity of CD.


Retrospective, single-center study including 146 patients with CD diagnosed between June 1983 and December 2011. We defined aggressive CD as need for more than 2 steroids courses required after diagnosis, need for surgery after diagnosis (except surgery for perianal disease or ileocecal resection as first choice of treatment) and need of admission for flare after diagnosis. The variables studied at diagnosis were smoking status, family history of inflammatory bowel disease, extraintestinal manifestations, Montreal Classification (age, location of CD, involvement of the upper gastrointestinal tract, phenotype, perianal disease), need for corticosteroids, admission and surgery on the first flare.

Statistical analysis was performed with SPSS vs 18.0 and a p value of less than 0.05 was considered statistically significant.


146 patients with CD were included, 55.5% female, with mean age 37.6±11.9 years and mean follow-up of 86±60 months. 80 patients presented at follow-up with non aggressive CD (55%) and 66 (45%) with aggressive CD. Independent factors present at diagnosis and significantly associated with aggressive CD were smoking status (p = 0.022), stenotic and penetrating phenotype (p < 0.001) and corticosteroids and admission on the first flare (p < 0.001). Given the results of the univariate analysis, in our cohort, the presence at diagnosis in an individual patient of 2 or more out of the 4 factors was associated with a high risk of aggressive disease with a accuracy of 0.73 (sensitivity of 88%, a specificity of 61%, a positive predictive value of 65% and a negative predictive value of 86%).


At diagnosis of Crohn's disease predictors of subsequent aggressive course are the initial requirement for corticosteroids and admission, smoking status and penetrating and stenotic disease. The combination of 2 or more variables is highly associated with aggressive disease and may be a useful way to make therapeutic decisions at diagnosis.