POS0288 A KEY TIF1γ EPITOPE MAY FACILITATE THE IDENTIFICATION OF PATIENTS AT HIGHEST RISK OF CANCER ASSOCIATED MYOSITIS

Abstract

Background:The myositis specific autoantibody anti-TIF1γ targets TRIM33, a TRIM family proetin with a PHDBromo domain at the C terminal end. Anti-TIF1γ is strongly associated with malignancy in adult patients with idiopathic inflammatory myopathies (IIM). Intriguingly, anti-TIF1γ is also the most common autoantibody in juvenile-onset IIM but younger patients with anti-TIF1γ do not have an increased risk of cancer (1-3). Genetic studies have consistently shown human leukocyte antigen (HLA) to be the strongest risk factor for IIM. Adult and juvenile-onset patients with anti-TIF1γ have recently been shown to have different associations at the HLA-DQB1 locus (4). This could be due to differences in the key TIF1γ epitopes and may relate to differences in aetiology, such as malignancy in adults versus other environmental factors in juvenile onset disease.Objectives:To identify key epitopes targeted by anti-TIF1γ antibody in patients with IIM and establish if different TIF1γ epitopes are targeted in patients with and without malignancy.Methods:Patient plasma/serum samples were obtained from UK Juvenile Dermatomyositis Cohort and Biomarker and UKMyoNet studies. Autoantibody status had previously been determined by immunoprecipitation. Cancer data was collated from the UK Health and Social Care Information Centre and cancer associated IIM (CAM) defined as that occurring within 3 years of IIM diagnosis.An in house ELISA was developed using a purified TIF1γ fragment comprising residues 882-1090, produced in E.coli, corresponding to the PHDBromo protein domain. An ELISA cut-off of 5SD above the mean of 38 healthy control samples was used.Results:38 healthy controls, and 117 anti-TIF1γ IIM patient sera (60 juvenile onset) were analysed for reactivity to the TIF1γ PHDbromo domain.No healthy controls were positive. Anti-TIF1γPHDbromo was more common in JDM: 18 (30%) juvenile patients and 6 (10.5%) adult patients were positive, p=0.01.Additional data was available for 39 adult patients (82% female, median age 52 (IQR 38-64)). Anti-TIF1γPHDbromo was only found in CAM patients plus one young adult non-CAM patient aged 27 years at disease onset, p =0.07 (p=0.02 for patients >30 years at disease onset), see Table 1. No juvenile patients had a history of malignancy.Table 1.Anti-TIF1γPHDbromo in 39 adult patients with cancer data availableanti-TIF1γPHDbromo positiveanti-TIF1γPHDbromo negativeCancer associated myositis n(%)5a (83)12(36)Myositis not associated with cancer n(%)1b(17)21(64)Total633a. Median age 64 yearsb. Age 27 yearsConclusion:The TIF1γ PHDBromo domain is an important epitope and autoantibody reactivity is more common in patients with juvenile-onset disease and adults with CAM. The distinction between adult and juvenile-onset IIM is arbitrary and it is relevant that the only adult with anti-TIF1γPHDbromo not to have malignancy was just 27 years old. Our numbers are small and further work is needed to establish if anti-TIF1γPHDbromo, and indeed other TIF1γ epitopes, could help identify patients at highest risk of malignancy, and how this relates to our understanding of IIM aetiopathogenesis. Our findings may add weight to the theory that the development of IIM in younger patients occurs after immune-mediated resolution of a pre-cancer event.