Predictive biomarkers in rheumatoid arthritis

Abstract

Being a very heterogeneous disease, rheumatoid arthritis (RA) is challenging for treatment. On a group level, some therapy options might be superior compared with others, however, this does not mutually exclude the less effective option to be more suitable for some patients, if the superior therapy option does not help. In different patients the concentration of biomarkers may vary dramatically, however, translation of meaning of these variations for each patient is not feasible yet. Observations and comparisons of these biomarkers before start of therapy between patient groups with different outcome after therapy can help to understand their role and make individualised approach for the therapy choice.

Low or moderate levels of a multi-biomarker disease activity (MBDA) score measured at baseline or follow-up visits could identify RA patients at very low risk of radiographic progression (RP). Moreover, in patients with high MBDA score at the start of treatment escalation, those on infliximab+methotrexate (IFX+MTX) therapy had significantly less RP than patients on non-biological triple therapy (TT) (papers I and II).

In treatment-naive, early RA, patients who failed respond to MTX were randomized to IFX+MTX or non-biological TT. The categories of the MBDA score at the time of randomisation were differentially associated with treatment outcome after 1 year for these two therapies. Patients with low MBDA score benefited more from TT, while those with high MBDA score responded better to IFX (paper III).

Furthermore, when the 12 component biomarkers of the MBDA score were analysed at baseline, four of those (paper IV) as well as two of 177 proteins retrieved from an affinity proteomic study (paper V) were associated with treatment outcome: low disease activity (LDA) and EULAR good response after 3 months of MTX therapy. Combination of these biomarkers within each study also showed improved prediction of treatment outcome.

In patients failing on MTX monotherapy who were randomised to addition of biological TNF inhibitor IFX, very low serum IFX (sIFX) level and anti-drug antibody (ADA)-positivity were associated with poorer outcome. Among baseline parameters, female gender predicted very low sIFX level and ADA-positivity at follow-ups, with similar trend for RF-positivity (paper VI).

In summary, using combination of serum biomarkers helps to predict and identify preferential therapy option for subsets of patients. Further studies of these biomarkers, if validated, will facilitate personalised therapy approach for subsets of patients.