Author(s): Oldroyd A.; Sergeant J.C.; Chinoy H.; New P.; Cooper R.; McHugh N.J.; Betteridge Z.; Lamb J.; Ollier W.
Source: Arthritis and Rheumatology; Oct 2017; vol. 69
Publication Date: Oct 2017
Publication Type(s): Conference Abstract
Abstract:Background/Purpose: There is an increased cancer risk associated with the idiopathic inflammatory myopathies (IIM). Studies have identified that positivity for the autoantibody against transcriptional intermediary factor 1 (anti-TIF-1 Ab) confers an even greater cancer risk in the IIMs. Investigating the temporal relationship between anti-TIF-1 Ab positivity and cancer onset in a large longitudinal IIM cohort will be inform future cancer screening practice. This study aimed to characterise the temporal relationship between anti-TIF-1 Ab positivity and cancer onset in a large UK-based adult IIM cohort. Methods: Data from adults with a Bohan & Peter-verified diagnosis of IIM from the UKMYONET study were analysed. Anti-TIF-1 Ab (alpha and gamma serosubtype) positive and negative patients were included. Each patient was followed up until they either developed cancer or were censored due to death. UKMYONET recruitment began in 1999, and cancer occurrence linkage was carried out up until December 2016 through the UK Health and Social Care Information Centre. Cancer associated myositis (CAM) was defined as an incident cancer occurring three years either side of the onset of IIM. The cumulative incidence of cancer after IIM onset was estimated according to Kaplan-Meier methods for the anti-TIF-1 Ab positive and negative cohorts. Hazard ratios for the time to cancer diagnosis by anti-TIF-1 Ab positivity were calculated using a Cox-regression model adjusted for age, gender and smoking status. Results: Data from 711 IIM cases were analysed, with a total of 8009 person-years follow up (Table 1); 55 (8%) of the IIM cases were anti-TIF-1 Ab positive, and all had dermatomyositis. A higher proportion of the anti-TIF-1 Ab positive cohort developed CAM, compared to the anti-TIF-1 Ab negative cohort: 38% vs 8%. Even after only one year of follow up, the proportion of the anti-TIF-1 Ab positive patients developing cancer (27%) exceeded the three year proportion of the anti- TIF-1 Ab negative cohort (8%). Cox proportional modelling revealed that anti-TIF-1 Ab positivity was significantly associated with an increased hazard of an associated cancer following IIM onset: HR 3.4 (95% CI 2.2, 5.4). Breast (33%), ovarian (19%) and lymphoma (14%) were the three most common cancers in the anti-TIF-1 Ab positive patients, whereas breast (20%), bowel (14%), lung (6%) and cervix (6%) were the most common sites in anti-TIF-1 Ab negative patients. Conclusion: This study has helped to characterise the temporal relationship between anti-TIF-1 Ab positivity and cancer onset. The findings that earlier cancer onset is associated with anti-TIF-1 Ab positivity, and that associated cancer types differ, are potentially of clinical significance, and appear to suggest a specific cancer screening approach in anti-TIF-1 Ab positive patients. (Table Presented).