Elsevier

Joint Bone Spine

Volume 81, Issue 4, July 2014, Pages 325-330
Joint Bone Spine

Original article
Gene expression profile predicting the response to anti-TNF treatment in patients with rheumatoid arthritis; analysis of GEO datasets

https://doi.org/10.1016/j.jbspin.2014.01.013Get rights and content

Abstract

Objectives

Anti-tumor necrosis factor (TNF) therapy is the treatment of choice for rheumatoid arthritis (RA) patients in whom standard disease-modifying anti-rheumatic drugs are ineffective. However, a substantial proportion of RA patients treated with anti-TNF agents do not show a significant clinical response. Therefore, biomarkers predicting response to anti-TNF agents are needed. Recently, gene expression profiling has been applied in research for developing such biomarkers.

Methods

We compared gene expression profiles reported by previous studies dealing with the responsiveness of anti-TNF therapy in RA patients and attempted to identify differentially expressed genes (DEGs) that discriminated between responders and non-responders to anti-TNF therapy. We used microarray datasets available at the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO).

Results

This analysis included 6 studies and 5 sets of microarray data that used peripheral blood samples for identification of DEGs predicting response to anti-TNF therapy. We found little overlap in the DEGs that were highly ranked in each study. Three DEGs including IL2RB, SH2D2A and G0S2 appeared in more than 1 study. In addition, a meta-analysis designed to increase statistical power found one DEG, G0S2 by the Fisher's method.

Conclusion

Our finding suggests the possibility that G0S2 plays as a biomarker to predict response to anti-TNF therapy in patients with rheumatoid arthritis. Further investigations based on larger studies are therefore needed to confirm the significance of G0S2 in predicting response to anti-TNF therapy.

Introduction

Rheumatoid arthritis (RA) is a systemic autoimmune disease characterized by chronic inflammation of joints followed by destruction of cartilage and bone. The precise etiology of RA is not completely understood, but a large body of studies indicates that pro-inflammatory cytokines such as TNFα and IL-1β are significantly involved in the pathophysiology of joint inflammation and bone destruction in RA [1]. Biologic agents, especially anti-TNF agents, have produced significant and clinically relevant improvements in patients with active RA and appear to be very effective therapeutic molecules for RA patients. Although the use of anti-TNF agents has revolutionized the treatment of RA, a substantial proportion (up to one-third) of RA patients treated with anti-TNF agents fail to achieve a satisfactory clinical response [2]. Therefore, a strategy for predicting which patients will and will not respond prior to initiation of therapy is needed to prevent disease progression, unnecessary side effects, and high costs for non-responders. Although baseline demographics and clinical characteristics, including age, sex, disease duration, Disease Activity Score (DAS28), Health Assessment Questionnaire score, and concurrent use of disease-modifying anti-rheumatic drugs (DMARDs), have been extensively studied, potential prediction markers of therapeutic responsiveness have not yet been discovered [3].

Pharmacogenetic studies have also been performed to discover determinants of efficacy of anti-TNF agents. Some studies have shown that polymorphisms in genes encoding TNFα, TNF receptors, the major histocompatibility complex region, and the Fcγ receptor IIIA (FCγRIIIA) are associated with a favorable response to anti-TNF therapy; however, these results were not consistent with those of other studies [3], [4].

Gene expression profiling with microarrays has become a standard method for identifying the genes and biological pathways that are associated with various complex diseases [5], [6]. Biomarkers identified by gene expression profiling can be used to make a diagnosis, monitor disease activity, and predict response to therapy. This technology has already been shown to identify gene expression profiles predictive of treatment response and clinical outcomes for some diseases [7], [8], [9], [10], [11]. Recently, several studies have attempted to identify gene expression profiles predicting the response to anti-TNF therapy in RA patients by using multiple gene expression microarrays [12], [13], [14], [15], [16], [17]. Although these studies have successfully identified gene expression signatures predicting the response to anti-TNF therapy, the expression profiles identified in these studies were not consistent with each other, and the genes reported by the authors of each study had little overlap. Herein, we attempt to identify differentially expressed genes (DEGs) that differ between responders and non-responders to anti-TNF therapy by:

  • comparing gene signatures previously reported to predict response to anti-TNF therapy in RA patients;

  • analyzing microarray datasets of these studies, which are available at the National Center for Biotechnology Information's Gene Expression Omnibus (GEO), using GEO2R;

  • performing a meta-analysis of these datasets.

Section snippets

Data collection

We performed literature survey via Medline/PubMed searches to identify available studies that examined gene expression profiles of peripheral blood samples to predict treatment outcome of anti-TNF therapy in RA patients. We searched for a combination of KEY words such as “Infliximab”, “Adalimumab”, “Etanercept”, “Microarray”, “Gene expression profile”, “Anti-TNF” and “Rheumatoid arthritis” as both medical subject heading terms and text words. We also tried to identify additional studies by

Study characteristics

The analyses included 6 studies that used peripheral blood samples to identify gene expression profiles predicting response to anti-TNF therapy. The dataset of 1 of these studies was not deposited in the GEO database [15]. Therefore, the microarray dataset of this study was not included for individual analysis or meta-analysis. Detailed information of each study is described in Table 1. In the study of Lequerré et al., patients with very active disease (Disease Activity Score 28 > 5.1) were

Discussion

We performed a comprehensive literature search to identify microarray-based studies of gene expression profiles predicting response to anti-TNF therapy in RA patients and then attempted to find a DEG profile to discriminate between different response groups by combining the datasets of these studies. To our knowledge, this is the first study in which publicly available microarray datasets for anti-TNF response were combined and re-examined by means of the same analytical tool such as GEO2R and

Disclosure of interest

The authors declare that they have no conflicts of interest concerning this article.

Acknowledgement

This research was supported by a Korea University Grant.

References (33)

  • B.L. Ebert et al.

    An erythroid differentiation signature predicts response to lenalidomide in myelodysplastic syndrome

    PLoS Med

    (2008)
  • A. Holleman et al.

    Gene-expression patterns in drug-resistant acute lymphoblastic leukemia cells and response to treatment

    N Engl J Med

    (2004)
  • S.L. Pomeroy et al.

    Prediction of central nervous system embryonal tumour outcome based on gene expression

    Nature

    (2002)
  • L.J. van’t Veer et al.

    Gene expression profiling predicts clinical outcome of breast cancer

    Nature

    (2002)
  • A. Julia et al.

    An eight-gene blood expression profile predicts the response to infliximab in rheumatoid arthritis

    PLoS One

    (2009)
  • T. Lequerre et al.

    Gene profiling in white blood cells predicts infliximab responsiveness in rheumatoid arthritis

    Arthritis Res Ther

    (2006)
  • Cited by (34)

    • ScRNA-seq identified the metabolic reprogramming of human colonic immune cells in different locations and disease states

      2022, Biochemical and Biophysical Research Communications
      Citation Excerpt :

      Since the patients with non-response to the anti-TNF therapy significantly upregulated the OSM gene expression [32], we subset the cluster of OSM+ inflammatory monocytes. The OSM+ inflammatory monocytes possessed a higher expression of G0S2 and PTGS2 (noted anti-TNF resistance genes [7,19]) and the genes associated with inflammation, such as IL1B, CXCL2, and NFKBIA (Fig. 4B). Consistently, the gene sets of inflammatory response and TNFA signaling were enriched in OSM+ inflammatory monocytes versus OSM− ones (Fig. 4C and D).

    • Predictors of treatment response in rheumatoid arthritis

      2019, Joint Bone Spine
      Citation Excerpt :

      Other studies further support the use of the whole-blood transcriptome for predicting the treatment response [36,37]. However, in an analysis of data from the public Gene Expression Omnibus (GEO) obtained from five independent transcriptome studies, the response to TNFα antagonists was predicted by a single gene, GOS2, encoding a protein involved in cell proliferation, apoptosis, inflammation, metabolism, and carcinogenesis [38]. Another study looked at the gene combinations identified in eight transcriptome studies done to identify genetic features predicting the response to infliximab in 374 patients with RA, of whom 191 (51.1%) failed to respond [39].

    • Predictors of treatment response in rheumatoid arthritis

      2019, Revue du Rhumatisme (Edition Francaise)
    • Comprehensive analysis of key lncRNAs in ischemic stroke

      2020, Mathematical Biosciences and Engineering
    View all citing articles on Scopus
    View full text