Editorial: The Multi‐Biomarker Disease Activity Test for Rheumatoid Arthritis: Is It a Valid Measure of Disease Activity?

Abstract

One of the greatest unmet needs in rheumatology is a biomarker that enables highly sensitive and reliable measurement of disease activity in patients with rheumatoid arthritis (RA). Witness the enthusiasm of the rheumatology community for Vectra DA, the latest multibiomarker disease activity (MBDA) test, which was developed and marketed by Crescendo Biosciences and is available commercially in the US as a send-out laboratory test. This MBDA test measures 12 serum protein biomarkers using multiplex enzyme immunoassays and applies an algorithm to summarize the biologic information into a single score. The test is performed at the Crescendo Biosciences laboratory in South San Francisco, which is certified by the Clinical Laboratory Improvement Amendments and licensed in California. Vectra DA is described by Crescendo Biosciences as "Everything you need to know: the advanced blood test that helps you and your doctor better understand your RA disease activity with a single score" (1).In this issue of Arthritis & Rheumatology, Fleischmann and colleagues present data that appear to reveal substantial flaws in the validity of this test (2). The authors report the findings of a post hoc analysis of the AMPLE (Abatacept versus Adalimumab Comparison in Biologic-Naive RA Subjects with Background Methotrexate) 2-year phase IIIb randomized controlled trial (3). The AMPLE trial enrolled RA patients with a disease duration of ,5 years, who had moderate disease activity despite methotrexate therapy, and who were naive to biologic therapies. The results demonstrate that after baseline, MBDA-defined categories of low disease activity, moderate disease activity, or high disease activity were not consistent with the cutoffs for validated clinical measures of disease activity, including the Disease Activity Score in 28 joints using the C-reactive protein level (DAS28-CRP), the Clinical Disease Activity Index (CDAI), and the Routine Assessment of Patient Index Data 3 (RAPID-3). There were no associations between the MBDA score and the DAS28-CRP, CDAI, or RAPID-3 in either treatment arm. At the end of the study, most patients had moderate or high disease activity based on the MBDA scores, whereas the majority of patients had low disease activity or remission based on the clinical measures.Perhaps the finding of moderate or high disease activity as determined by the MBDA test despite clinical indicators of minimal disease activity could yet be meaningful. Vectra DA is marketed as being a "deeper" or more "comprehensive" measure of disease activity, suggesting that the MBDA test can detect disease persistence even when clinical measures are insensitive to it (4). However, the foundation of this idea appears to crack further because of data showing not only that MBDA-classified disease activity failed to predict radiographic outcomes but also that the highest probability of radiographic nonprogression was evident in the groups with MBDA-defined moderate or high disease activity. Considering the extensive efforts to bring the MBDA test to t...

Reply

“…In the absence of a clear understanding of which are the most relevant biomarkers to consider, we have looked at multiple types of biomarker analyses that could be useful in this endeavor; this included the MBDA test. We concur with the accompanying editorial published in the same issue of Arthritis & Rheumatology that encourages further efforts in this area (7), as there is a need to develop clinically relevant tools both for patients with RA and for clinicians. We look forward to the results of these efforts.…”

Reanalysis of the Multi‐Biomarker Disease Activity Score for Assessing Disease Activity in the Abatacept Versus Adalimumab Comparison in Biologic‐Naive Rheumatoid Arthritis Subjects with Background Methotrexate Study: Comment on the Article by Fleischmann et al

“…Thresholds for DAS28-CRP RA disease activity categories are systematically lower than the corresponding cutoffs in the DAS28 using the erythrocyte sedimentation rate (ESR) (15,16), as noted in the editorial accompanying Fleischmann and colleagues' article (17). The comparison of MBDA scores versus DAS28-CRP in the article used the DAS28-ESR category thresholds, and therefore yielded more discordance than would have been expected between MBDA categories and DAS28-CRP categories using DAS28-CRP thresholds.…”

Biomarkers in Rheumatoid Arthritis

“…There was a modest correlation between MBDA, DAS28-CRP, and DAS28-ESR, with weaker correlations observed with SDAI, Clinical Disease Activity Index (CDAI), and Routine Assessment of Patient Index Data (RAPID3) [ 42 ]. However, subsequent post hoc analysis of data from the AMPLE trial (abatacept versus adalimumab for RA) showed disagreement between the MBDA test score and these measures [ 43 - 44 ]. One trial showed that the MBDA test may be useful in deciding whether or not to continue biologic therapy in the setting of clinical remission [ 45 ], and post hoc analyses of data have shown that a high baseline MBDA score is a strong independent predictor of radiographic progression at one year [ 46 - 49 ].…”