Background Rheumatoid arthritis (RA) can be characterized by joint tissue destruction. TCZ treatments are effective in treating signs and symptoms, but only a subset of patients are protected from further joint destruction. Thus personalized medicine may be needed to identify those patients which will be global responders to level potentially side effects and cost of treatments. During joint destruction, small fragments of tissue are released into serum, where they may be quantified as serological biomarkers, termed a protein fingerprint.
Objectives To investigate whether biomarker changes after 1 dosing could identify subjects experiencing an early response, warranting continued treatment and confidence in treatment success.
Methods Biomarkers were measured in serum of 403 RA patients treated with either 4 or 8mg/kg TCZ (the LITHE study); C1M, C2M and C3M (MMP-degraded type I, II and III collagen), CRPM (MMP-degraded CRP), CTx/OC (bone balance), MMP3 and CRP. Early responders were patients who achieved ≥20% improvement in the swollen joint count and the tender joint count by week 16. The predictive power of the biomarkers, at baseline and %>change after first dose, for identification of responders and non-responders, was investigated by AUROC, logistic regression and CART analysis. The 8 and 4 mg/kg doses were investigated as two independent data sets and corrected to age, gender, and disease duration.
Results There was a significant difference in the suppression of biomarkers when comparing the two doses, illustrated by CRP level; CRP was completely suppressed by the 8 mg/kg dose, whereas 4 mg/kg suppressed CRP level by 30%. Different suppression levels were observed for early responders and non-responders when treated with 8 mg/kg: Baseline CTx/OC (AUC0.66, p=0.0006) and changes in C1M (AUC0.71, p=0.0010), C2M (AUC0.69, p=0.0092) and C3M (AUC0.68, p=0.014). A combination of the markers provided an AUC0.79 (p=0.0075). This was further supported by CART analysis where patients with either a high bone turnover (CTX/OC) or a low C1M and C2M were 6.7-fold more likely to be responders to TCZ (p<0.0001). 8 mg/kg equally suppressed CRP in responders and non-responders, whereas 4 mg/kg was separated by 20%, indicating that a full 8 mg/kg dose is needed for provide complete response to TCZ. This was further supported by measurement of the other markers, where little or no separation was seen between early responders and non-responders.
Conclusions We found that 8 mg/kg TCZ provided the best pharmacodynamic response on biomarkers of the joint tissue degradation, in comparison to 4mg/kg. This enhanced pharmacodynamic response enhanced identification of early responders, with a superior clinical benefit of TCZ treatment. The result of this biomarkers analysis may assist in the identification of the patients, who respond most optimally to given interventions, and potentially provide stronger risk/benefit and cost/value propositions to patients and payers.
AC Bay-Jensen. Effect of tocilizumab combined with methotrexate on circulating biomarkers of synovium, cartilage, and bone in the LITHE study. Semin Arthritis Rheum (2013).
JS Smolen. Tocilizumab inhibits progression of joint damage in rheumatoid arthritis irrespective of its anti-inflammatory effects: disassociation of the link between inflammation and destruction. Ann Rheum Dis (2012).
Acknowledgements Thank you to the LITHE patients and investigators.
Disclosure of Interest : A.-C. Bay-Jensen Grant/research support: FP7grants, Employee of: Nordic Bioscience, A. Platt Employee of: Former employee of Roche and a current employee of AstraZeneca, A. S. Siebuhr Grant/research support: PhD scholarship, C. Christiansen Shareholder of: Nordic Bioscience, M. Karsdal Shareholder of: Nordic Bioscience, Grant/research support: FP7grants, Employee of: Nordic Bioscience