Article Text
Abstract
Introduction Rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA) are important diagnostic tools in rheumatoid arthritis (RA). These antibodies are predominantly of the IgM (RF) or IgG (ACPA) isotype. Other isotypes of both antibodies – like IgA – and other antibodies – like anti-RA33, which is directed to the nuclear antigen hnRNP-A2 – have been repeatedly reported but their diagnostic and prognostic value has still not been fully elucidated.
Objectives Here we investigated (i) the prevalence of IgA-RF and IgA-ACPA as well as isotypes IgA, IgG and IgM of anti-RA33 antibody in patients with RA and (ii) their predictive value regarding therapeutic response to methotrexate (MTX).
Methods Sera from 290 RA patients, 261 disease controls and 100 healthy subjects were tested for the presence of RF, ACPA and anti-RA33 IgG/A/M isotypes by EliATM (Thermo Fisher Scientific). RF and ACPA were routinely measured by nephelometry and the anti-CCP EliATM, respectively. For finding associations with American College of Rheumatology (ACR)20 and simplified disease activity score (SDAI)50 therapeutic responses, an inception cohort of 165 RA patients was analysed.
Results Diagnostic specificity of antibodies was at least 95%. 185 (63.4%) of 290 RA patients tested positive for at least one routine marker (RF or ACPA) while 107 were negative for both antibodies (seronegative). Among these, 24 (8.2%) patients tested positive for IgG/A/M anti-RA33 and/or IgA-RF/ACPA. To determine the prognostic value regarding therapeutic responses a cross-validated combined model with an accuracy of 77% and an estimated p-value (k=10) of 0.00034 showed high levels (>133 IU/ml) of IgM-RF to be associated with a favourable response to methotrexate (MTX). In case of low or no RF, the presence of IgG-RA33 antibody on the one hand, and the absence of IgA-ACPA on the other hand was associated with a favourable response.
Conclusions Thus, these data suggest that determination of multiple antibodies increases the diagnostic power of serological testing and may be a feasible tool for the prediction of MTX response especially in combined models.
Disclosure of interest D. Sieghart: None declared, A. Platzer: None declared, F. Alasti: None declared, P. Studenic: None declared, M. Grundhuber Employee of: Thermo Fisher Scientific – Phadia GmbH, S. Swiniarski Employee of: Thermo Fisher Scientific – Phadia GmbH, S. Blueml: None declared, T. Perkmann: None declared, J. Smolen: None declared, G. Steiner: None declared