P705 Long-term outcomes and predictors of disabling disease of a population-based cohort of incident Crohn’s disease diagnosed between 1995 and 1997
M.-L. Rabilloud*1, J.-F. Bretagne1, E. Bajeux2, L. Siproudhis3, J.-F. Viel2, I. Tron4, M. Robaszkiewicz5, G. Bouguen1
1Université Rennes 1 & CHU Pontchaillou Rennes, Service des Maladies de l’Appareil Digestif, Rennes, France, 2CHU Pontchaillou, Service de Santé Publique, Rennes, France, 3Université Rennes 1 & CHU Pontchaillou Rennes, Service de chirurgie digestive, Rennes, France, 4ORSB, Rennes, France, 5CHU Brest, La Cavale Blanche, Service des Maladies de l’Appareil Digestif, Brest, France
Assessment of CD outcomes and identification of early prognostic factors that predict a severe course remain needed for physician decision-making to minimise structural bowel damage, complications of disease, and side effects related to treatment. Only population-based studies are able to assess these questions given the selection bias of referral centre studies
All incident cases of patients diagnosed with possible CD (n = 370) were registered from 1994 to 1997 in Brittany, a limited area in France. At diagnosis were recorded clinical features, endoscopic lesion per ileocolic segment (according to the CDEIS), and radiologic and histologic data. All charts of patients were reviewed from the diagnosis to the last clinic in 2015. Disabling Crohn’s disease course was defined according to the Saint-Antoine criteria: more than 2 steroid course or steroid dependency/resistance, hospitalisation related to CD, use of immunosuppressant or anti-TNF, and anal or major abdominal surgery.
After a median follow-up of 12.8 years, disease locations were broadly similar for each patient over time involving the ileon for 15% of patient, the colon for 36% of patient, and ileocolic for 48% of patients. Cumulative probabilities of stricturing disease and fistulising disease at 15 years were 35% and 17%, respectively. Cumulative probabilities of complicated behaviour (structuring or fistulising disease) were at 5 years, 10 years, and 15 years, 23%, 34%, and 42%, respectively. Cumulative probabilities of major abdominal surgery were at 1 year, 5 years, 10 years, and 15 years (14%, 29%, 36%, and 45%, respectively). The cumulative probabilities of the use of steroids, immunosuppressant (thiopurine and methotrexate) and TNF antagonist were 66% (n = 179), 37% (n = 102), and 22% (n = 60) at 15 years. Median disease duration before the use of immunosuppressant or TNF antagonist were 68 months and 131 months, respectively. Cumulative probabilities of disabling disease were at 1 year, 5 years, 10 years, and 15 years (35%, 57%, 67%, and 74%, respectively. Systemic manifestation at diagnosis (HR = 1.62, p = 0.02) and perianal lesion (HR = 4.47, p = 0,001) independently predicted a disabling disease course. Endoscopic lesions at diagnosis were not associated with disease outcomes.
On a population-based cohort, most of the patients experienced disabling disease outcomes according to the predefined Saint-Antoine criteria. The proportion of patient treated with immunosuppressant and/or TNF antagonist remained low. The high rate of disabling outcomes may underlined the need to treat CD patients with effective treatment whatever the disease characteristics at diagnosis, including endoscopic lesions.