P443 Risk of incident Cancer in Patients with Inflammatory Bowel Disease with Prior Breast Cancer: a Multicentre Cohort Study

Background Breast cancer is the most common malignancy observed in patients with inflammatory bowel diseases (IBD) unrelated to the disease or its treatment (Poullenot F et al. JCC 2022). The main aim of our study was to assess the risk of incident cancer according to the IBD treatment given in patients with prior breast cancer. Methods Consecutive IBD patients with prior breast cancer diagnosis were included in a multicenter retrospective cohort from 25 tertiary centres. Inclusion date corresponded to the diagnosis of index cancer. Follow up was calculated from the first administration of immunomodulator after cancer diagnosis (or cancer diagnosis date in absence of treatment) to the occurrence of incident cancer, corresponding to recurrence of breast cancer or de novo cancer, or to the last follow-up visit. Patients were categorized according to the use or not of immunomodulator after cancer diagnosis: thiopurines, methotrexate, anti-TNF, vedolizumab, ustekinumab. Crude incidence rates were compared between patients receiving at least one immunomodulator and those without immunomodulator before and after matching on age, lymph node, and metastasis extension and tumor’s grade, using a propensity-score analysis with a 1:1 ratio. Results Among the 151 identified patients, 80 patients with full available data were analyzed: 76 (95%) women; mean age at index cancer diagnosis: 51.5 years [standard deviation (SD): 11.5 years]; 44 (55%) with Crohn’s disease, 35 (44 %) ulcerative colitis and 1 (1 %) indeterminate colitis; median IBD duration at inclusion was 13 years [interquartile range (IQR) 6-21]. After a median follow up of 84 months [IQR 49-154], 16 (20%) incident cancers were observed: 12 (15%) recurrences and 4 (5%) cancer de novo. Three (4%) patients died from cancer related cause during the follow up. 39 (49%) patients received no immunomodulator and, 11 (14%) were treated with thiopurines, 6 (7.5%) with methotrexate, 18 (22.5%) with anti-TNF, 5 (6%) with vedolizumab and one (1%) with ustekinumab. Those treatments were initiated with a median interval of 24 months [IQR 7-48] after cancer diagnosis. Crude incidence rate per 1000 person-years were 47.97 for patients not exposed to any immunomodulator and 12.61 for the others (p=0.0248). After matching, adjusted crude incidence rates per 1000 person-years were 50 and 27.27, respectively (p=0.3798). Rates of survival without incident cancer were not different between the two groups after matching on age, lym