P704 Safety of Immunosuppression in A Prospective Cohort of Inflammatory Bowel Disease Patients with a HIstoRy of CancEr: The SAPPHIRE Registry

For patients with inflammatory bowel disease (IBD) and a history of cancer, retrospective studies have failed to find evidence suggesting that exposure to biologic and/or immunomodulator (IMM) agents was associated with new or recurrent cancer. The SAPPHIRE Registry was developed to prospectively examine this issue. Patients with IBD who had a confirmed first (index) cancer 5 years prior to enrollment were recruited from centers affiliated with the New York Crohn’s and Colitis Organization (NYCCO) since 2016. Patients receiving chemotherapy or radiation therapy at enrollment or recurrent cancer within the last five years were excluded. The primary outcome was the development of an incident cancer stratified by IBD therapy exposure and adjusted for index cancer stage, index cancer type, sex, smoking history, and age at index cancer in a Cox proportional hazards model. 302 patients were analyzed. Median age at IBD diagnosis was 31.5. Patients were 47% male, 88% white, and 61% never smokers. Index cancers were solid organ (138; 46%), dermatologic (100; 33%) GI (34; 11%), hematologic (27; 9%), or of other types (3; 1%); most index cancers (excluding non-melanoma skin cancer) were diagnosed as Stage 1. Following the index cancer, 90 (30%) patients were not exposed to any immunosuppression, and the other 212 were exposed to immunosuppressive therapy: 142 biologic monotherapy (anti-TNF, anti-integrin, anti-IL12/23, anti-IL23; 47%), 13 anti-metabolite monotherapy (mercaptopurine, azathioprine, methotrexate; 4%), 41 biologic + anti-metabolite combination therapy (14%), and 16 exposed to small molecules (JAK inhibitors, ozanimod; 5%). During follow-up, 46 (15%) patients developed subsequent cancers (25 new, 21 recurrent): 22 (48%) dermatologic, 15 (33%) solid, 7 (15%) GI, and 2 (4%) hematologic malignancies at a median age of 63.5. Patients not exposed to immunosuppression experienced incident cancer at a rate of 2.58/100 person-years (PY, Table). Those exposed to immunosuppression had a rate of 5.12/100 PY, representing an additional 2.54 cancers/100 PY (95% CI: 0.22, 4.85). In a time-varying Cox proportional hazards model adjusting for clinical and demographic characteristics, the adjusted hazard ratio for incident cancer for patients exposed to immunosuppressive therapy compared to unexposed patients was 1.36 (95% CI: 0.73, 2.52). Excluding non-melanoma skin cancer, the adjusted hazard ratio was 1.38 (95% CI: 0.66, 2.91). In this ongoing prospective study, we have thus far found no association between immunosuppressive IBD therapy and risk of developing new or recurrent cancer. Ongoing enrollment and follow-up, specifically for individual drug classes and cancer types, are required to confirm these findings.