P652. Early treatment with immunomodulators is associated with change in the natural history of inflammatory bowel disease – multicentre longitudinal cohort study – Sydney, Australia
V. Kariyawasam1, T. Huang2, P. Lunney2, K. Middleton2, R. Wang2, C. Selinger2, P. Katelaris2, J. Andrews3, R. Leong2, 1Guy's and St Thomas' NHS Foundation Trust, Gastroenterology, Londond, United Kingdom, 2Concord Repatriation General Hospital, Gastroenterology, Concord, Australia, 3Royal Adelaide Hospital, Dept of Gastroenterology & Hepatology, Adelaide, Australia
Immunomodulators (azathioprine, mercaptopurine and methotrexate) (IM) are effective in the treatment of Crohn's disease (CD) and ulcerative colitis (UC) but few population studies have had sufficient power to demonstrate their impact on natural history of disease.
Sydney IBD Cohort was first published in 1995. Longitudinal follow up of this cohort plus further enrichment of new cases identified using identical methodologies were included. Patient demographics, disease characteristics according to Montreal classification, drug and surgical treatments were reviewed. Comparisons were made between 3 time periods of pre 1990 (A), 1990–2000 (B) and post 2000 (C). Kaplan–Meier, log rank and Cox proportional hazard regression model statistics were used.
A total of 1118 CD patients were included with a median follow-up of 11 years and 14530 patient years of follow-up. Median age of diagnosis was 29 years. The locations of disease was L1 24, L2 37 and L3 39%, disease behaviour was B1 73, B2 18 and B3 9% and perianal involvement was 18% at baseline. Mean time to commence IM use was 15.9, 6.0 and 1.3 years in the three time periods (P < 0.0001). Cumulative probability of commencing IM was significantly reduced over the said time periods (P < 0.0001). Long-term steroid (LTS) use (P < 0.0001), surgical resection and recurrent resection (P = 0.002) also decreased significantly. Use of IM early on in the disease course prior to any complications predicted for decreased surgery and progression of disease (B1 towards B2 and/ or B3) behaviour (P < 0.0001).
A total of 1171 UC patients were recruited with a median follow-up of 12 years and 18,035 patient years of follow-up. The median age of diagnosis was 33 years and the extent of disease being E1 33, E2 41 and E3 26% at baseline. Time to introduction of IM significantly decreased over successive decades (P < 0.0001) and was associated with a significant reduction in the use of LTS (P < 0.0001). Surgical resection rates also significantly decreased during the follow up period (P = 0.033, A vs B&C: P = 0.012). Use of IM within 2 years of diagnosis predicted the decreased need for colectomy (P = 0.012).
The study demonstrates a significantly increased probability of early introduction of immunomodulators over time for both Crohn's disease and ulcerative colitis. Early introduction of IM was found to be associated with reduced progression of disease behaviour in CD and reduced requirement for surgery in both CD and UC.