THU0201 Ability of the laboratory tests to predict rheumatoid arthritis in patients with early arthritis. (1st June 2001)

• Record Type: Journal Article
• Authors: Saraux, A; Berthelot, JM; Chales, G; Le Henaff, C; Mary, JY; Thorel, JB; Hoang, S; Dueymes, M; Devauchelle, V; Baron, D; Le Goff, P; Youinou, P
• Is Part Of: Annals of the rheumatic diseases. Volume 60(2001)Supplement 1
• Page Start: A435
• Page End: A435
• Publication Date: 2001-06-01
• Subjects: Rheumatism -- Periodicals; 616.723005
• DOI: 10.1136/annrheumdis-2001.1103
• Languages: English
• ISSNs: 0003-4967

Abstract

Abstract : Background: Objectives: To determine the diagnostic value in early rheumatoid arthritis (RA) of various laboratory tests used alone or in combination. Methods: A cohort of 270 patients with early arthritis was constituted between 1995 and 1997 in seven hospitals in Brittany (France). Each patients underwent at first visit a standardised examination, laboratory tests [rheumatoid factors (RF), the functionnal affinity of IgM-rheumatoid factors, antiperinuclear factor, IgG, IgM and IgA antikeratin antibody (AKA), antiRA33 antibody, antinuclear antibodies, ELISA analysis of IgG glycosylation, circulating immune complexes, IgG 1, IgG2, IgG3, and IgG 4, and HLA AB DR typing], and radiographs. They were examined in the same way at six-month intervals until their last visit between June and December 1999, when the diagnosis of RA was made by a panel of five rheumatologists. Ninety-eight (38%) patients were considered as having RA at their final visit. The ability of each test as measured at the initial visit to discriminate between RA and non RA patients was evaluated through sensitivity and specificity. Optimal cutoff for continuous test was derived from Received Operating Characteristic curve. Multiple logistic regression was used to select combination of laboratory tests able to discriminate between RA and non RA patients and the diagnostic value of these combination was evaluated through sensitivity and specificity. Results: The performance of the combinations using either the pair with the best diagnostic value combined with a third selected laboratory test or multiple logistic regression demonstrated that high levels of IgM RF by ELISA, IgG AKA, and latex test had the strongest association to RA. Conclusion: IgM-RF by ELISA, IgG-AKA, and the latex test are the best laboratory tests for discriminating between patients with and without RA. The other biological tests did not improve the diagnostic value.