P262. Clinical outcome of perianal Crohn's disease: Natural history and impact of medical and surgical strategies over time
A. Natalis1, E. Louis1, C. Vankemseke1, J. Belaiche1, L. Seidel2, C. Reenaers1, 1CHU de Liege, Gastroenterologie, Liege, Belgium, 2CHU de Liege, Statistics, Liege, Belgium
Perianal Crohn's disease (pCD) is associated with complications leading to recurrent surgery and tissue damage. Immunosuppressive drugs (IS) including anti-TNF have changed the management of pCD. Our aim was to describe the management and the natural history of a cohort of patients with active pCD and to identify predictive factors of poor evolution.
A retrospective study of pCD patients registred in the database of the university hospital of Liège, Belgium. Perianal lesions included abscess, fistulae, anal fissure, anal strictures. pCD treatments included antibiotics, surgical drainage (with or without seton), stoma. Medical treatments including IS and anti-TNF were recorded at pCD diagnosis and over follow-up. pCD relapse was defined as antibiotherapy for recurrent abscess, the need for surgical drainage or stoma. The subroups of patients followed before (old cohort) and after (young cohort) the year 2000 were compared in a subanalysis.
181 patients with pCD were included. Mean follow-up was 7.9 years Mean time between CD and pCD diagnosis was 6.3 years. Lesions at pCD diagnosis were abscess in 93/181 (51%), fistula in 91/181 (50%; 77/93 of complex fistulae), anal fissure in 28/181 (15%), anal stricture in 18/181 (10%). At diagnosis abscess drainage was performed in 31/181 (17%), drainage + seton in 44/181 (24%), stoma in 18/181 (10%). 132/181 (74%) and 83/181 (47%) had IS and anti-TNF respectively at pCD diagnosis. Relapse rate was 51% within a mean time of 33 months. During follow-up 15% required a stoma. Predictive factors of relapse were perianal abscess (p < 0.0001, HR=4.4), fistula (p < 0.0001, HR=4.5) or surgical drainage at diagnosis (p < 0.0001, HR=4.5), young age at pCD diagnosis (28 versus 31 yo, p = 0.02), short time between CD and pCD diagnosis (5.7 versus 7 years, p = 0.01), IS (p = 0.04, HR=1.8) and anti-TNF (p = 0.01, HR=1.5) at pCD diagnosis. Anti-TNF during follow-up, time to introduce them and duration of anti-TNF treatment were not predictive of relapse. The young and old cohort had the same characteristics at pCD diagnosis except a higher use of IS (87% vs 48%, p < 0.0001) and anti-TNF (3% vs 68%, p < 0.0001) in the young cohort. Clinical outcome including the time to relapse, type of relapse, need for surgery and stoma was similar in both cohorts.
In our cohort of pCD patients half of them had a perianal relapse over the time requiring surgery in more than 2/3 of them. At pCD diagnosis perianal abscess, fistula, surgical drainage, young age, treatment with IS or anti-TNF were associated with a higher risk of relapse. Although higher prescription of anti-TNF and IS in the last years new treatment strategies have not impacted the outcome of pCD.