FRI0007 Prediction of Therapeutic Responses to TOCILIZUMAB in Patients with Rheumatoid Arthritis Using Biomarkers Identified by Genome-Wide DNA Microarray Analysis in Peripheral Blood Mononuclear Cells

Background Although the overall response rate of rheumatoid arthritis (RA) to tocilizumab is high, improvement of synovitis in patients treated with tocilizumab is frequently slow as compared with that in patients treated with TNF antagonists. Therefore, predicting therapeutic responses of RA to tocilizumab can be particularly beneficial but such methods with sufficient accuracy have not been established yet. Objectives This prospective multicenter study aimed to identify biomarkers to predict therapeutic responses to tocilizumab in patients with RA by analyzing comprehensive gene expression in peripheral blood mononuclear cells (PBMCs). Methods Patients with RA who received treatment with tocilizumab for the first time were recruited and therapeutic responses were determined at 6 months. In the training cohort (n=40), gene expression in PBMCs at baseline was analyzed using genome-wide DNA microarray with 41,000 probes representing 19,416 genes. In the validation cohort (n=20), expression levels of the candidate genes in PBMCs at baseline were determined using real-time quantitative PCR (qPCR) analysis. Results In hierarchical clustering analyses, gene expression patterns of non-responders clustered in the same branch (Figure) and 68 DNA microarray probes which showed a significant difference in signal intensity between non-responders and responders in the training cohort were identified. Nineteen putative genes were selected and a significant correlation between the DNA microarray signal intensity and the qPCR relative expression was confirmed for 15 genes. In the validation cohort, a significant difference in relative expression between non-responders and responders was reproduced for 3 type I interferon response genes (IFI6, MX2, OASL) and MT1G. Models incorporating these genes provided a maximum area under the receiver operating characteristic curve of 0.947 in predicting a moderate or good response in the validation cohort. Conclusions Our study using genome-wide DNA microarray analyses identified the candidate biomarkers to predict therapeutic responses to tocilizumab in patients with RA and suggests the involvement of type I interferon signaling and metallothioneins in the pathophysiology of RA. Acknowledgements This work was supported by a Health Labour Sciences Research Grant on Allergic Disease and Immunology from the Ministry of Health, Labor and Welfare of Japan, and a Regional Innovation Strategy Support Program and a Grants-in-Aids for Scient