P081 Malignancy rates in children, young people, and adults with juvenile idiopathic arthritis using UK primary care data (CPRD Aurum)

Abstract Background/Aims Juvenile idiopathic arthritis (JIA) affects ∼1-2/2’000 children and young peoplein the UK. In adults with rheumatoid arthritis, there is ∼10% increasedmalignancy risk compared with the general population. Malignancy rates inNordic children with JIA range from 3-5/10’000 person years, with conflictingevidence on risk versus the general population. The aim of this research is tocalculate the malignancy rate of patients with JIA in England, compared with(a) matched-controls, and (b) general population estimates. Methods Using UK primary care data (CPRD Aurum), this analysis matched patients with JIA (prevalent and incident) 16 years old, 4-to-1 with non-JIA controls based on birth year, gender, and practice. Malignancies were identified through (i) NHS-linked hospitalisation data (all malignancies with an ICD-10 code starting with ‘C’), and (ii) CPRD read codes using a pre-defined code list. Exposure started on first JIA code date (or matched-date for controls), 1-Jan-2000, or CPRD entry date, whichever was latest. Follow-up continued until end of follow-up of JIA-matched patient (for control patients only), end of CPRD follow-up, cut-off date (31-Dec-2018), first reported malignancy, or death, whichever was first. Cox-proportional hazards model was used to compare malignancy rates in JIA versus matched-controls. Standardised incidence ratios (SIRs) were generated to compare rates with the ONS general population estimates, based on calendar year, year of age, and gender. All patients were from England. Results 3,751 people with JIA and 10,864 matched controls were identified; characteristics were similar between cohorts (Table). Malignancy rates were 6.8 per 10,000 (95%CI 4.3-10.6) for JIA patients and 4.9 per 10,000 person years (95% CI 3.6 to 6.8) for control patients (HR = 1.36; 95%CI 0.78-2.37). Both SIRs were raised compared with ONS general population estimates; 3.3 for JIA (95% CI 2.1-5.1), and 2.6 for controls (95% CI 1.9-3.6). Conclusion This analysis calculated malignancy rates in young people with JIA using linked primary and secondary care records. No significant difference in malignancy rates were identified between the JIA and control cohort. However, both SIRs were higher compared with what was expected considering their age and gender indicating that rates in this population (CPRD) are likely an overestimation versus national malignancy rates. Disclosure L. Kearsley-Fleet: None. S. Merriel: None. N. Shaw: None. J. Leslie: None. M. Johnson: None. L.R. Wedderburn: None. K.L. Hyrich: None. J.H. Humphreys: None.