Background/Purpose:  There are limited data regarding the role of and response to treatment, of bone biomarkers in early rheumatoid arthritis (RA) treated with conventional DMARDs. We sought to determine whether levels of bone biomarkers associated with osteoclast activation [RANKL and Dickkopf-1 (Dkk-1)] or inhibition [osteoprotegerin (OPG)] correlated with response to treatment in an inception cohort of RA patients receiving treat-to-target combination DMARD therapy.

Methods:  Patients with early RA (< 1 year; fulfilling 2010 classification criteria, n=112) received triple therapy (methotrexate, sulfasalazine and hydroxychloroquine) escalated to achieve DAS28 remission, without oral corticosteroids. RANKL, OPG and Dkk-1 were analysed by Luminex kits at 0, 6 and 12 month and in healthy controls (n=33). OPG levels were log-transformed prior to analysis; as a significant proportion of RANKL levels were below detection, results were dichotomised (positive/negative). Correlations between biomarkers and changes following treatment were analysed using Spearman’s rank coefficient and mixed-model longitudinal regression respectively.

Results: At baseline, 69% were positive (‘seropositive’) for rheumatoid factor (RF) and/or anti-CCP (cyclic-citrullinated peptide), mean (±SD) age was 58(13) years, 72% were females, 60% current/past smokers, mean symptom duration prior to diagnosis of 18 (±11) weeks and mean DAS28 was 5.52 (1.30). Median (IQR) total SvH score was 3 (8); 20% had erosive disease. At baseline, compared to controls, OPG levels were elevated in both seronegative (p <0.001) and seropositive (p <0.001) patients. In seropositive patients, RANKL was more frequently detectable (p < 0.001) and negatively correlated with Dkk-1 (p <0.05). In contrast, seronegative patients had higher Dkk-1 (p <0.001) which strongly correlated with OPG (p <0.001) and RANKL was not significantly different from controls (p =0.36). Following treatment, mean DAS28 at 6 and 12 months was 3.55 (1.55) and 3.28 (1.60) respectively. In seropositive patients, there was a significant reduction in proportion of patients with detectable RANKL (p =0.002) and an increase in OPG levels (p <0.01) but no significant change in Dkk-1 levels. Seronegative patients, in contrast, had no change in proportion of patients with detectable RANKL (p=0.76) or OPG levels (p=0.5), but had a significant reduction in Dkk-1 (p <0.001).

Conclusion: Individual bone biomarkers show significant differences at baseline and following conventional DMARD treatment, between seropositive and seronegative patients. Reduction in Dkk-1 and reduction in RANKL and increase in OPG may be useful biomarkers of treatment response in seronegative and seropositive patients respectively. The differential response of biomarkers, depending on seropositivity, may help direct therapeutic decisions in order to optimise prevention of joint damage. Further analysis of these biomarkers in relation to disease activity and erosive disease is warranted.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/seropositivity-predicts-bone-biomarker-change-in-an-inception-cohort-of-rheumatoid-arthritis-patients-treated-to-target-with-combination-conventional-dmard-therapy/