A MULTI-BIOMARKER BASED DISEASE ACTIVITY (MBDA) SCORE SYSTEM COMPARED TO A CONVENTIONAL DISEASE ACTIVITY SCORE (DAS) SYSTEM IN THE BEST RHEUMATOID ARTHRITIS (RA) STUDY

Background: Although the outcome of rheumatoid arthritis (RA) patients is better when care is guided by regular measurement of a disease activity (DAS) score, in some circumstances measurement of DAS is not carried out, due to lack of access to a rheumatologist or limited resources. A multi-biomarker based disease activity (MBDA) score could provide an objective, precise information about disease activity which would help rheumatologists for a simpler management of the RA patients. Objectives: To confirm the relationship between MBDA scores and conventional DAS methods including DAS28, DAS28CRP, and DAS44, to validate the MBDA algorithm. Additionally, to provide evidence-based cut off values for the use of the MBDA as a guide for management of RA patients in clinical practice. Methods: A total of 124 RA patients from the BeSt clinical trial were studied. Clinical data and serum samples were available from 180 visits, 91 at baseline and 89 at year 1. Clinical assessments were carried out and calculated DAS44, DAS28, and DAS28CRP. The MBDA algorithm combines serum biomarkers (VCAM-1, EGF, VEGF-A, IL-6, TNF-RI, MMP-1, MMP-3, YKL-40, Leptin, Resistin, CRP, SAA) into a single score and is used in Vectra™ DA, a novel validated multi-biomarker blood test for RA disease activity which is commercially available in the US. The 12 biomarkers were measured in each serum sample using quantitative multiplex immunoassays, and the concentrations were used as input to the MBDA algorithm to calculate score. The association between MBDA and conventional DAS (DAS28, DAS28CRP, and DAS44) and the association between ΔMBDA and ΔDAS28 were evaluated by Pearson's correlation. The values of MBDA stratified by EULAR disease activity (LDA, MDA, or HDA) were compared by one-way factorial ANOVA. Linear regression with MBDA as predictor and DAS28 as response was used to define an MBDA cutoff equivalent to DAS28 =3.2. A receiver operating characteristic (ROC) analysis was developed with MBDA as predictor and DAS28 <3.2 as response. The association of categorical variables was assessed by Fisher's exact test. Results: MBDA scores had a significant correlation with DAS28, DAS28CRP, and DAS44 (p<0.0001 in all cases). The cumulative probability plot of MBDA showed similar distribution to those of conventional DAS values. Groups stratified by EULAR disease activity (DAS28 <3.2, 3.2 to 5.1, and >5.1), were significantly different (p<0.0001) and there was relatively little overlap between the high and low disease activity groups. The MBDA score could discriminate low disease activity (DAS28 <3.2) with area under ROC curve of 0.83 (P<0.0001), and score threshold of 28 or less had a >90% specificity, whereas score of 44 or higher had a >90% negative predictive value. Changes in MBDA were also correlated to changes in DAS28 (cor =0.54, p<0.0001). Conclusions: MBDA is a useful surrogate for conventional disease activity measurement. The cut-off values of 28 and 44 may be recommended to alter the therapeutic strategy for RA. Disclosure of Interest: S. Hirata: None Declared, L. Dirven: None Declared, Y. Shen Employee of: Crescendo Bioscience Inc., M. Centola: None Declared, G. Cavet Employee of: Crescendo Bioscience Inc., W. Lems: None Declared, Y. Tanaka Consultant for: Mitsubishi Tanabe Pharma, Takeda Pharmaceutical Co Ltd, Abbott, Eisai Pharma, T. Huizinga Consultant for: Schering Plough, UCB, Bristol Myers Squibb, Biotest AG, Wyeth/Pfizer, Novartis, Roche, Sanofi-Aventis, Abbott, Crescendo Bioscience, Axis-Shield diagnostics, C. Allaart: None DeclaredCitation: Annals of the Rheumatic Diseases, volume 70, supplement 3, year 2011, page 593Session: Rheumatoid arthritis – comorbidity and clinical aspects (Poster Presentations )