Agueusop Inoncent, Margerie Daniel, Remaury Anne, Brard Raphaël, Frau Francesca, Gerard Emilie, Thill Gilbert, Veeranagouda Yaligara, Didier Michel, Kohlmann Markus, Herrmann Matthias, Biesemann Nadine
Sanofi R&D, Biostatistics, Industriepark Hoechst, Frankfurt am Main, Germany.
Sanofi R&D, Digital and Data Science, Industriepark Hoechst, Frankfurt am Main, Germany.
RMD Open. 2025 Aug 4;11(3):e005556. doi: 10.1136/rmdopen-2025-005556.
To identify blood-based predictive and pharmacodynamic biomarkers at different timepoints in patients with active rheumatoid arthritis (RA) treated with anti-interleukin-6 receptor (anti-IL-6R) and anti-tumour necrosis factor-α (anti-TNF-α).
This study used blood samples from the MONARCH trial (NCT02332590), a randomised, double-blind, phase III trial that compared the safety and efficacy of sarilumab (anti-IL-6R) and adalimumab (anti-TNF-α) monotherapy in patients with RA who were intolerant/inadequate responders to methotrexate. The study evaluated predictive biomarkers to anti-IL-6R and anti-TNF-α treatments at baseline and week 2 and pharmacodynamic biomarkers at week 2 and week 24 using Olink proteomics analysis (n=804 serum samples from 268 patients). Change in gene expression levels (n=522 peripheral blood samples from 261 patients) by both treatments was assessed using RNA sequencing analysis.
Serum biomarkers most predictive to anti-IL-6R were different from those of anti-TNF-α; predictive biomarkers for anti-IL-6R were correlated with innate immune activation and synovial inflammation, while predictive biomarkers for anti-TNF-α seemed to be more T-cell and neutrophil-related. For baseline predictive biomarkers, we had to focus on relative prediction as the absolute prediction performance of single and combination biomarkers using cross-validation was limited. Additionally, the pharmacodynamic effects of anti-IL-6R and anti-TNF-α on biomarkers as well as pathway signatures were distinct.
The unbiased analysis of serum proteins identified biomarkers most predictive of anti-IL-6R and anti-TNF-α at different timepoints that could explain the difference in the response rate in patients with RA. Further, both biomarker and pathway results highlighted a differentiated mode of action of both treatments.
识别接受抗白细胞介素-6受体（抗IL-6R）和抗肿瘤坏死因子-α（抗TNF-α）治疗的活动性类风湿关节炎（RA）患者在不同时间点基于血液的预测性生物标志物和药效学生物标志物。
本研究使用了MONARCH试验（NCT02332590）的血样，这是一项随机、双盲、III期试验，比较了sarilumab（抗IL-6R）和阿达木单抗（抗TNF-α）单药治疗对甲氨蝶呤不耐受/反应不足的RA患者的安全性和有效性。该研究使用Olink蛋白质组学分析（来自268名患者的804份血清样本）评估了基线和第2周时对抗IL-6R和抗TNF-α治疗的预测性生物标志物，以及第2周和第24周时的药效学生物标志物。使用RNA测序分析评估了两种治疗方法引起的基因表达水平变化（来自261名患者的522份外周血样本）。
对抗IL-6R最具预测性的血清生物标志物与抗TNF-α的不同；抗IL-6R的预测性生物标志物与固有免疫激活和滑膜炎症相关，而抗TNF-α的预测性生物标志物似乎更多与T细胞和中性粒细胞相关。对于基线预测性生物标志物，由于使用交叉验证的单一和组合生物标志物的绝对预测性能有限，我们不得不关注相对预测。此外，抗IL-6R和抗TNF-α对生物标志物以及通路特征的药效学作用是不同的。
对血清蛋白的无偏分析确定了在不同时间点对抗IL-6R和抗TNF-α最具预测性的生物标志物，这些生物标志物可以解释RA患者反应率的差异。此外，生物标志物和通路结果都突出了两种治疗方法不同的作用模式。
