Email updates

Keep up to date with the latest news and content from Arthritis Research & Therapy and BioMed Central.

This article is part of the supplement: 23rd European Workshop for Rheumatology Research

Meeting abstract

Rheumatoid arthritis – class prediction by autoreactivity profiles

R Bergholz1, F Schumann1, S Behrens1, U Ungethüm1, G Valet2, WA Schmidt3, GR Burmester1, JM Engel4, WJ van Venrooij5, G Steiner6 and S Bläβ1

Author Affiliations

1 Department of Rheumatology & Clinical Immunology, Charité University Clinic, Berlin, Germany

2 MPI Biochemistry, Munich, Germany

3 Clinic for Rheumatology Berlin Buch, Berlin, Germany

4 Rheumaklinik, Bad Liebenwerda, Germany

5 Department of Biochemistry, University of Nijmegen, The Netherlands

6 Divison of Rhematology, Department Internal Medicine III, Vienna General Hospital, Austria

For all author emails, please log on.

Arthritis Res Ther 2003, 5(Suppl 1):1  doi:10.1186/ar631


The electronic version of this article is the complete one and can be found online at:


Published:24 February 2003

©

Meeting abstract

Heterogeneity and multifactoriality complicate diagnostics and our understanding of pathogenesis of rheumatoid arthritis (RA). The only accepted serologic parameter (rheumatoid factor [RF]) is not disease specific, nor are any of several novel RA autoantibodies. We aimed at identifying profiles instead of individual autoreactivities allowing for unambiguous prediction of RA.

Selected RA autoantigens were tested by ELISA (RF and anti-cyclic citrullinated peptide [anti-CCP]) or Western blot (heavy-chain-binding protein [BiP], heterogeneous ribonucleoprotein particle A2 [RA33/hnRNP A2], calpastatin and calreticulin). Antibody reactivities were assayed from serum samples of 149 RA patients and 132 patients with other rheumatic diseases and from synovial fluids (SF) (58 RA, 65 non-RA).

No single autoreactivity was sufficient for unambiguous prediction of RA. Frequencies of multiparameter profiles consisting of 3, 4, 5 and 6 autoreactivites were determined. Fifteen six-parameter serum profiles were exclusively expressed in RA patients, representing a cumulative sensitivity of 59%. Twelve SF profiles were exclusively expressed in 64% of RA patients. The self-learning classification algorithm CLASSIF1 was capable of accurately predicting RA when these profiles were present. Data profile analysis of RF/CCP/BiP/calpastatin/calreticulin/RA33 provided specific discrimination of 64% of RA. Most importantly, RA specific profiles were observed in 64% of patients with early disease (<12 months).

For the first time, the accurate prediction of the class RA has been achieved by the use of multiparametric autoreactivity profiles. Because of early expression in disease, these profiles make it possible to start a disease-modifying therapy long before irreversible bone and joint destruction may develop. Additional RA-specific profiles are required to cover the entire group of RA patients.