Distinct HLA associations with autoantibody-defined subgroups in idiopathic inflammatory myopathies

Background In patients with idiopathic inflammatory myopathies (IIM), autoantibodies are associated with specific clinical phenotypes suggesting a pathogenic role of adaptive immunity. We explored if autoantibody profiles are associated with specific HLA genetic variants and clinical manifestations in IIM. Methods We included 1348 IIM patients and determined the occurrence of 14 myositis-specific or-associated autoantibodies. We used unsupervised cluster analysis to identify autoantibody-defined subgroups and logistic regression to estimate associations with clinical manifestations, HLA-DRB1, HLA-DQA1, HLA-DQB1 alleles, and amino acids imputed from genetic information of HLA class II and I molecules. Findings We identified eight subgroups with the following dominant autoantibodies: anti-Ro52, -U1RNP, -PM/Scl,-Mi2,-Jo1,-Jo1/Ro52,-TIF1 gamma or negative for all analysed autoantibodies. Associations with HLA-DRB1*11, HLA-DRB1*15, HLA-DQA1*03, and HLA-DQB1*03 were present in the anti-U1RNP-dominated subgroup. HLA-DRB1*03, HLA-DQA1*05, and HLA-DQB1*02 alleles were overrepresented in the anti-PM/Scl and anti-Jo1/ Ro52-dominated subgroups. HLA-DRB1*16, HLA-DRB1*07 alleles were most frequent in anti-Mi2 and HLA- DRB1*01 and HLA-DRB1*07 alleles in the anti-TIF1 gamma subgroup. The HLA-DRB1*13, HLA-DQA1*01 and HLA-DQB1*06 alleles were overrepresented in the negative subgroup. Significant signals from variations in class I molecules were detected in the subgroups dominated by anti-Mi2, anti-Jo1/Ro52, anti-TIF1 gamma, and the negative subgroup. Interpretation Distinct HLA class II and I associations were observed for almost all autoantibody-defined subgroups. The associations support autoantibody profiles use for classifying IIM which would likely reflect underlying pathogenic mechanisms better than classifications based on clinical symptoms and/or histopathological features. Funding See a detailed list of funding bodies in the Acknowledgements section at the end of the manuscript. Copyright (c) 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).