DNA labyrinth leading to personalized RA treatment.

RA Response Predictors: Can Biomarkers Guide Your Treatment?

Lena Kashyap in Health & Wellness December 2025 • 4 min read.

"Unlock personalized RA therapy: Discover how predictive biomarkers like ACPA, transcriptomes, and proteomes are transforming treatment decisions."

The increasing number of available treatments for rheumatoid arthritis (RA) presents a challenge: choosing the right molecule for each patient at a specific time. After the failure of a disease-modifying antirheumatic drug (DMARD), rheumatologists now have access to a range of therapeutic options. These options include eleven treatments belonging to six distinct therapeutic classes, not counting biosimilars, each targeting different pathophysiological mechanisms.

Clinical remission is achieved in only about 10% to 30% of patients across studies, with variable efficacy at the individual patient level. These factors are encouraging rheumatologists to seek clinical and biological markers to predict treatment response in a personalized medicine approach.

Predicting treatment response involves anticipating how a patient will respond to a specific treatment before it is administered. This approach helps avoid exposing patients to ineffective molecules, saving time, reducing costs, and improving the benefit-risk balance. Prediction is based on the pathophysiological, clinical, and therapeutic heterogeneity of RA. Studies have shown varied profiles of cellular infiltrations, cytokine expressions, and gene expressions in RA synovial tissue, even when patients initially have similar clinical characteristics.

Unlocking Personalized RA Treatment: How Biomarkers Predict Response

DNA labyrinth leading to personalized RA treatment.

Identifying and validating biomarkers to predict response to RA treatments requires measuring these markers in patients who respond to treatments versus those who do not. Response to treatment becomes a criterion that needs definition. Should response be defined using EULAR criteria, remission criteria, CDAI criteria, ACR criteria, or structural response? Although structural response remains the ultimate goal, it is challenging to assess because structural evolution varies among patients with the same activity level and progression rate.

Rheumatologists also face primary non-response (0 to 6 months) or instances of treatment escape, which can occur quickly or later, regardless of the molecule used. Furthermore, response measurement timelines vary from 3 months to 6 months, or even longer. Response is not binary; van Vollenhoven RF et al. showed in 406 RA patients treated with etanercept or infliximab that response to anti-TNF-α agents followed a normal, sometimes asymmetrical distribution, but never a bimodal one, suggesting that biological mechanisms determining treatment response are multifactorial.

Clinical and Biological Markers: Initial attempts focused on clinical and biological markers used for RA diagnosis or prognosis, but these markers do not reliably predict treatment response.
Antibodies: While rheumatoid factors (RF) and anti-citrullinated protein antibodies (ACPA) are useful for RA diagnosis and associated with response to rituximab or abatacept, a meta-analysis showed they do not predict response to anti-TNF-α biologics.
Matrices: Matrices that combine clinical and biological parameters show promise for personalized treatment approaches.

Matrices, similar to the SCORE system in cardiology, integrate several parameters simultaneously for individual-level results. For example, in 3280 patients treated with golimumab, six parameters (male sex, young age, low HAQ, CRP, ESR, number of tender or swollen joints, and absence of comorbidity) were identified to predict remission.

The Future of RA Treatment: Personalized Approaches

The number of theranostic biomarkers remains limited. Diagnostic, prognostic, and pathophysiological biomarkers are generally not discriminating enough, with the exception of RF and anti-CCP for certain molecules like rituximab and abatacept. Matrices using these parameters offer an innovative research direction for individual patients. Approaches without a priori, based on transcriptomic and proteomic tools, appear most promising for identifying reliable biomarkers. However, most identified biomarkers have not been replicated due to methodological heterogeneity and small sample sizes. Integrative biology, currently growing, should enable the discovery of combinations capable of predicting response to different treatments. Personalized medicine remains a challenge to address in the coming years due to the increasing number of molecules.

About this Article -

This article was crafted using a human-AI hybrid and collaborative approach. AI assisted our team with initial drafting, research insights, identifying key questions, and image generation. Our human editors guided topic selection, defined the angle, structured the content, ensured factual accuracy and relevance, refined the tone, and conducted thorough editing to deliver helpful, high-quality information.See our About page for more information.

This article is based on research published under:

DOI-LINK: 10.1016/j.rhum.2018.03.007, Alternate LINK

Title: Facteurs Prédictifs De Réponses Dans La Polyarthrite Rhumatoïde

Subject: Rheumatology

Journal: Revue du Rhumatisme

Publisher: Elsevier BV

Authors: Thierry Lequerré, Pascal Rottenberg, Céline Derambure, Pascal Cosette, Olivier Vittecoq

Published: 2019-01-01

Everything You Need To Know

What is the primary challenge in treating Rheumatoid Arthritis (RA), and how do biomarkers aim to address it?

The central challenge in treating Rheumatoid Arthritis (RA) is selecting the most effective treatment from a growing number of available options. Currently, rheumatologists can choose from eleven treatments across six therapeutic classes, excluding biosimilars. The variability in patient response, with clinical remission achieved in only 10% to 30% of cases, underscores the need for a personalized approach. Biomarkers, such as ACPA, transcriptomes, and proteomes, are being investigated to predict treatment response, helping to match patients with the most suitable medication, thereby improving outcomes, reducing exposure to ineffective treatments, and optimizing the benefit-risk balance.

How can biomarkers, specifically ACPA, transcriptomes, and proteomes, help in predicting the response to RA treatments, and what are the implications of using these biomarkers?

Predicting treatment response with biomarkers involves identifying measurable indicators that correlate with a patient's reaction to a specific treatment before it's administered. For example, the presence of anti-citrullinated protein antibodies (ACPA) has been associated with a response to rituximab or abatacept. Moreover, transcriptomes and proteomes, which provide insights into gene expression and protein profiles, respectively, are showing promise in identifying reliable biomarkers. The implications of using these biomarkers are significant: they enable personalized medicine by guiding the selection of the most effective treatment, avoiding the use of ineffective medications, reducing costs, and improving patient outcomes. This approach is particularly important given the heterogeneous nature of RA and the varied efficacy of existing treatments.

Why aren't clinical and biological markers, like RF and ACPA, always reliable in predicting the response to anti-TNF-α biologics?

While clinical and biological markers such as rheumatoid factors (RF) and anti-citrullinated protein antibodies (ACPA) are valuable for RA diagnosis and can predict response to certain treatments like rituximab and abatacept, they don't reliably predict response to anti-TNF-α biologics. A meta-analysis highlighted this limitation, indicating that the mechanisms determining response to these treatments are multifactorial and more complex than what these individual markers can reveal. This underscores the need for more comprehensive approaches, such as the use of matrices that combine multiple parameters, and the exploration of transcriptomic and proteomic tools to identify additional reliable biomarkers.

What are 'matrices' in the context of RA treatment, and how do they improve the prediction of treatment response?

In the context of Rheumatoid Arthritis (RA) treatment, 'matrices' refer to systems that integrate multiple clinical and biological parameters to predict individual patient responses. Similar to the SCORE system in cardiology, these matrices analyze several factors simultaneously to provide a comprehensive assessment. An example is the identification of six parameters (male sex, young age, low HAQ, CRP, ESR, number of tender or swollen joints, and absence of comorbidity) to predict remission in patients treated with golimumab. By considering multiple variables, matrices offer a more nuanced and accurate prediction of treatment response compared to relying on single biomarkers.

What are the main obstacles in the development of personalized RA treatment, and what strategies are being pursued to overcome them?

The primary obstacles in developing personalized RA treatment include the limited number of theranostic biomarkers and the methodological heterogeneity and small sample sizes in research. Diagnostic, prognostic, and pathophysiological biomarkers are often not discriminating enough, except in specific cases like RF and anti-CCP for treatments like rituximab and abatacept. To overcome these challenges, researchers are focusing on several strategies: developing matrices that combine various clinical and biological parameters, and employing 'approaches without a priori', based on transcriptomic and proteomic tools, to identify more reliable biomarkers. Integrative biology is also gaining traction, aiming to discover combinations of biomarkers that can effectively predict responses to different treatments. The increasing number of available molecules further complicates the treatment landscape, making personalized medicine a crucial and ongoing area of research.