Résumé
To determine the frequency and the predictive factors of colectomy and coloproctectomy in patients with ulcerative colitis. We conducted an 11-year retrospective study based on hospitalized ulcerative colitis patients followed up for more than 6 months. From 1995 to 2005, 115 patients were included [50 men, 65 women, mean age: 38.4 years]. Mean duration of follow-up was 39.2 months [6 - 145]. Colectomy was performed in 20 patients [17%], with an actuarial risk of 16% at 5 years and 35% at 10 years. Proctocolectomy with ileoanal anastomosis was performed in 16 cases and total colectomy with ileorectal anastomosis in 4 cases. In univariate analysis, factors associated with an increased risk of colectomy were pancolitic location [p=0.02], acute severe colitis [p < 0.0001], treatment by intravenous corticosteroids [p < 0.0001] and intravenous cyclosporine [p=0.001]. In multivariate analysis, acute severe colitis was the only independent factor associated with colectomy [p = 0.04 OR[CI95%] :6.66[1.04 - 50]]. In patients with distal location, the independent factor associated with colectomy was colonic extension during follow up [p = 0.04 OR[CI95%] :7.69[1.07 - 50]]. In patients with pancolitic location, risk of colectomy was associated with acute severe colitis [p = 0.01 OR[CI95%] :9.09[1.58 - 50]] and years of hospitalization from 1995 to 1999 [p = 0.02 OR[CI95%] :7.14[1.35 - 44]]. Although the diffusion of treatment by intravenous cyclosporin, surgery is frequently performed in our ulcerative colitis patients, specially in case of acute severe colitis. Evaluation of colonic extension during the follow-up is associated with an important prognostic impact