Nguyen Minh Vu Chuong, Courtier Anaïs, Adrait Annie, Defendi Federica, Couté Yohann, Baillet Athan, Guigue Lisa, Gottenberg Jacques-Eric, Dumestre-Pérard Chantal, Brun Virginie, Gaudin Philippe
GREPI EA 7408, Université Grenoble Alpes, 38000, Grenoble, France.
Sinnovial, 38000, Grenoble, France.
Clin Rheumatol. 2020 Sep;39(9):2553-2562. doi: 10.1007/s10067-020-05030-6. Epub 2020 Mar 24.
Rheumatoid arthritis (RA) is a debilitating disease, but patient management and treatment have been revolutionized since the advent of bDMARDs. However, about one third of RA patients do not respond to specific bDMARD treatment without clear identified reasons. Different bDMARDs must be tried until the right drug is found. Here, we sought to identify a predictive protein signature to stratify patient responsiveness to rituximab (RTX) among patients with an insufficient response to a first anti-TNFα treatment.
Serum samples were collected at baseline before RTX initiation. A proteomics study comparing responders and nonresponders was conducted to identify and select potential predictive biomarkers whose concentration was measured by quantitative assays. Logistic regression was performed to determine the best biomarker combination to predict good or nonresponse to RTX (EULAR criteria after 6 months' treatment).
Eleven biomarkers potentially discriminating between responders and nonresponders were selected following discovery proteomics. Quantitative immunoassays and univariate statistical analysis showed that fetuin-A and thyroxine binding globulin (TBG) presented a good capacity to discriminate between patient groups. A logistic regression analysis revealed that the combination of fetuin-A plus TBG could accurately predict a patient's responsiveness to RTX with an AUC of 0.86, sensitivity of 80%, and a specificity of 79%.
In RA patients for whom a first anti-TNFα treatment has failed, the serum abundance of fetuin-A and TBG before initiating RTX treatment is an indicator for their response status at 6 months. ClinicalTrials.gov identifier: NCT01000441. Key Points • Proteomic analysis revealed 11 putative predictive biomarkers to discriminate rituximab responder vs. nonresponder RA patients. • Fetuin-A and TBG are significantly differentially expressed at baseline in rituximab responder vs. nonresponder RA patients. • Algorithm combining fetuin-A and TBG accurately predicts response to rituximab in RA patients with insufficient response to TNFi.
类风湿性关节炎（RA）是一种使人衰弱的疾病，但自从生物改善病情抗风湿药（bDMARDs）问世以来，患者管理和治疗发生了革命性变化。然而，约三分之一的RA患者对特定的bDMARD治疗无反应，且原因不明。必须尝试不同的bDMARDs，直到找到合适的药物。在此，我们试图确定一种预测性蛋白质特征，以在对首次抗TNFα治疗反应不足的患者中分层对利妥昔单抗（RTX）的反应性。
在开始RTX治疗前的基线期收集血清样本。进行了一项蛋白质组学研究，比较反应者和无反应者，以鉴定和选择潜在的预测生物标志物，其浓度通过定量测定法测量。进行逻辑回归以确定预测对RTX良好反应或无反应的最佳生物标志物组合（治疗6个月后的欧洲抗风湿病联盟标准）。
经过发现蛋白质组学后，选择了11种可能区分反应者和无反应者的生物标志物。定量免疫测定和单变量统计分析表明，胎球蛋白-A和甲状腺素结合球蛋白（TBG）具有很好的区分患者组的能力。逻辑回归分析显示，胎球蛋白-A加TBG的组合可以准确预测患者对RTX的反应性，曲线下面积（AUC）为0.86，敏感性为80%，特异性为79%。
在首次抗TNFα治疗失败的RA患者中，开始RTX治疗前血清中胎球蛋白-A和TBG的丰度是其6个月反应状态的指标。ClinicalTrials.gov标识符：NCT01000441。要点•蛋白质组学分析揭示了11种推定的预测生物标志物，以区分利妥昔单抗反应者与无反应者的RA患者。•在利妥昔单抗反应者与无反应者的RA患者中，胎球蛋白-A和TBG在基线时显著差异表达。•结合胎球蛋白-A和TBG的算法可准确预测对TNF抑制剂反应不足的RA患者对利妥昔单抗的反应。
