PT - JOURNAL ARTICLE AU - Moder, Kevin G AU - Wener, Mark H AU - Weisman, Michael H AU - Ishimori, Mariko L AU - Wallace, Daniel J AU - Buckeridge, David L AU - Homburger, Henry A TI - Measurement of antinuclear antibodies by multiplex immunoassay: a prospective, multicenter clinical evaluation. DP - 2007 May 01 TA - The Journal of Rheumatology PG - 978--986 VI - 34 IP - 5 4099 - http://www.jrheum.org/content/34/5/978.short 4100 - http://www.jrheum.org/content/34/5/978.full SO - J Rheumatol2007 May 01; 34 AB - OBJECTIVE: We conducted a prospective, multicenter evaluation of autoantibody testing by multiplex immunoassay in patients with known or suspected connective tissue diseases (CTD). We evaluated agreement between multiplex immunoassay and enzyme immunoassay (EIA) and assessed the diagnostic utility of autoantibody profiles. METHODS: Samples from 908 patients with suspected CTD seen in rheumatology clinics were collected prospectively at 3 tertiary care centers. Diagnoses were established according to recognized classification criteria. Tests for autoantibodies were obtained by multiplex immunoassay and by EIA. The results of the multiplex immunoassay were analyzed using a previously validated interpretative algorithm, MDSS (Medical Decision Support Software), that suggests possible disease associations based on the pattern of results for the autoantibodies. RESULTS: The median patient age was 49.7 years; 83% were female. The most common diagnoses were rheumatoid arthritis in 352 patients and systemic lupus erythematosus (SLE) in 332 patients. Agreement between multiplex and EIA testing ranged from a high of 99% (95% CI 98% to 100%) for Jo-1 to a low of 79% (95% CI 76% to 82%) for antinuclear antibodies. The MDSS algorithm suggested an appropriate disease association in 75% to 100% of patients with SLE. The results varied depending on the disease and the autoantibodies present. CONCLUSION: These results suggest that patterns of autoantibodies detected by multiplex immunoassay testing, when analyzed by an interpretative algorithm, are useful in the evaluation of patients with CTD in situations of high disease prevalence. Further testing is necessary to determine its utility in settings of low disease prevalence.