Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis: A retrospective observational study

Seiji Nakamura, Katsuya Suzuki, Hiroshi Iijima, Yuko Hata, Chun Ren Lim, Yohei Ishizawa, Hideto Kameda, Koichi Amano, Kenichi Matsubara, Ryo Matoba, Tsutomu Takeuchi

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Background: According to EULAR recommendations, biologic DMARDs (bDMARDs) such as tumor necrosis factor inhibitor, tocilizumab (TCZ), and abatacept (ABT) are in parallel when prescribing to rheumatoid arthritis (RA) patients who have shown insufficient response to conventional synthetic DMARDs. However, most prediction studies of therapeutic response to bDMARDs using gene expression profiles were focused on a single bDMARD, and consideration of the results from the perspective of RA pathophysiology was insufficient. The aim of this study was to identify the specific molecular biological features predicting the therapeutic outcomes of three bDMARDs (infliximab [IFX], TCZ, and ABT) by studying blood gene expression signatures of patients before biologic treatment in a unified test platform. Methods: RA patients who responded inadequately to methotrexate and were later commenced on any one of IFX (n = 140), TCZ (n = 38), or ABT (n = 31) as their first biologic between May 2007 and November 2011 were enrolled. Whole-blood gene expression data were obtained before biologic administration. Patients were categorized into remission (REM) and nonremission (NON-REM) groups according to CDAI at 6 months of biologic therapy. We employed Gene Set Enrichment Analysis (GSEA) to identify functional gene sets differentially expressed between these two groups for each biologic. Then, we compiled "signature scores" for these gene sets, and the prediction performances were assessed. Results: GSEA showed that inflammasome genes were significantly upregulated with IFX in the NON-REM group compared with the REM group. With TCZ in the REM group, B-cell-specifically expressed genes were upregulated. RNA elongation, apoptosis-related, and NK-cell-specifically expressed genes were upregulated with ABT in the NON-REM group. Logistic regression analyses showed that "signature scores" of inflammasomes, B-cell-specifically expressed, and NK-cell-specifically expressed genes were significant, independently predictive factors for treatment outcome with IFX, TCZ, and ABT, respectively. The AUCs of ROC curves of these signature scores were 0.637, 0.796, and 0.768 for IFX, TCZ, and ABT, respectively. Conclusions: We have identified original gene expression predictive signatures uniquely underlying the therapeutic effects of IFX, TCZ, and ABT. This is, to our knowledge, the first attempt to predict therapeutic effects of three drugs concomitantly using a unified gene expression test platform.

Original languageEnglish
Article number159
JournalArthritis Research and Therapy
Volume18
Issue number1
DOIs
Publication statusPublished - 2016 Jul 19

Fingerprint

Biological Factors
Transcriptome
Observational Studies
Rheumatoid Arthritis
Retrospective Studies
Antirheumatic Agents
Genes
Inflammasomes
Therapeutic Uses
Therapeutics
Natural Killer Cells
B-Lymphocytes
Gene Expression
Biological Therapy
tocilizumab
Abatacept
Methotrexate
ROC Curve
Area Under Curve
Infliximab

Keywords

  • BDMARDs (biologic agents)
  • Gene expression
  • Prediction
  • Rheumatoid arthritis

ASJC Scopus subject areas

  • Rheumatology
  • Immunology
  • Immunology and Allergy

Cite this

Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis : A retrospective observational study. / Nakamura, Seiji; Suzuki, Katsuya; Iijima, Hiroshi; Hata, Yuko; Lim, Chun Ren; Ishizawa, Yohei; Kameda, Hideto; Amano, Koichi; Matsubara, Kenichi; Matoba, Ryo; Takeuchi, Tsutomu.

In: Arthritis Research and Therapy, Vol. 18, No. 1, 159, 19.07.2016.

Research output: Contribution to journalArticle

Nakamura, Seiji ; Suzuki, Katsuya ; Iijima, Hiroshi ; Hata, Yuko ; Lim, Chun Ren ; Ishizawa, Yohei ; Kameda, Hideto ; Amano, Koichi ; Matsubara, Kenichi ; Matoba, Ryo ; Takeuchi, Tsutomu. / Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis : A retrospective observational study. In: Arthritis Research and Therapy. 2016 ; Vol. 18, No. 1.
@article{af68b55a7ec748ccac334686262a37e5,
title = "Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis: A retrospective observational study",
abstract = "Background: According to EULAR recommendations, biologic DMARDs (bDMARDs) such as tumor necrosis factor inhibitor, tocilizumab (TCZ), and abatacept (ABT) are in parallel when prescribing to rheumatoid arthritis (RA) patients who have shown insufficient response to conventional synthetic DMARDs. However, most prediction studies of therapeutic response to bDMARDs using gene expression profiles were focused on a single bDMARD, and consideration of the results from the perspective of RA pathophysiology was insufficient. The aim of this study was to identify the specific molecular biological features predicting the therapeutic outcomes of three bDMARDs (infliximab [IFX], TCZ, and ABT) by studying blood gene expression signatures of patients before biologic treatment in a unified test platform. Methods: RA patients who responded inadequately to methotrexate and were later commenced on any one of IFX (n = 140), TCZ (n = 38), or ABT (n = 31) as their first biologic between May 2007 and November 2011 were enrolled. Whole-blood gene expression data were obtained before biologic administration. Patients were categorized into remission (REM) and nonremission (NON-REM) groups according to CDAI at 6 months of biologic therapy. We employed Gene Set Enrichment Analysis (GSEA) to identify functional gene sets differentially expressed between these two groups for each biologic. Then, we compiled {"}signature scores{"} for these gene sets, and the prediction performances were assessed. Results: GSEA showed that inflammasome genes were significantly upregulated with IFX in the NON-REM group compared with the REM group. With TCZ in the REM group, B-cell-specifically expressed genes were upregulated. RNA elongation, apoptosis-related, and NK-cell-specifically expressed genes were upregulated with ABT in the NON-REM group. Logistic regression analyses showed that {"}signature scores{"} of inflammasomes, B-cell-specifically expressed, and NK-cell-specifically expressed genes were significant, independently predictive factors for treatment outcome with IFX, TCZ, and ABT, respectively. The AUCs of ROC curves of these signature scores were 0.637, 0.796, and 0.768 for IFX, TCZ, and ABT, respectively. Conclusions: We have identified original gene expression predictive signatures uniquely underlying the therapeutic effects of IFX, TCZ, and ABT. This is, to our knowledge, the first attempt to predict therapeutic effects of three drugs concomitantly using a unified gene expression test platform.",
keywords = "BDMARDs (biologic agents), Gene expression, Prediction, Rheumatoid arthritis",
author = "Seiji Nakamura and Katsuya Suzuki and Hiroshi Iijima and Yuko Hata and Lim, {Chun Ren} and Yohei Ishizawa and Hideto Kameda and Koichi Amano and Kenichi Matsubara and Ryo Matoba and Tsutomu Takeuchi",
year = "2016",
month = "7",
day = "19",
doi = "10.1186/s13075-016-1052-8",
language = "English",
volume = "18",
journal = "Arthritis Research and Therapy",
issn = "1478-6354",
publisher = "BioMed Central",
number = "1",

}

TY - JOUR

T1 - Identification of baseline gene expression signatures predicting therapeutic responses to three biologic agents in rheumatoid arthritis

T2 - A retrospective observational study

AU - Nakamura, Seiji

AU - Suzuki, Katsuya

AU - Iijima, Hiroshi

AU - Hata, Yuko

AU - Lim, Chun Ren

AU - Ishizawa, Yohei

AU - Kameda, Hideto

AU - Amano, Koichi

AU - Matsubara, Kenichi

AU - Matoba, Ryo

AU - Takeuchi, Tsutomu

PY - 2016/7/19

Y1 - 2016/7/19

N2 - Background: According to EULAR recommendations, biologic DMARDs (bDMARDs) such as tumor necrosis factor inhibitor, tocilizumab (TCZ), and abatacept (ABT) are in parallel when prescribing to rheumatoid arthritis (RA) patients who have shown insufficient response to conventional synthetic DMARDs. However, most prediction studies of therapeutic response to bDMARDs using gene expression profiles were focused on a single bDMARD, and consideration of the results from the perspective of RA pathophysiology was insufficient. The aim of this study was to identify the specific molecular biological features predicting the therapeutic outcomes of three bDMARDs (infliximab [IFX], TCZ, and ABT) by studying blood gene expression signatures of patients before biologic treatment in a unified test platform. Methods: RA patients who responded inadequately to methotrexate and were later commenced on any one of IFX (n = 140), TCZ (n = 38), or ABT (n = 31) as their first biologic between May 2007 and November 2011 were enrolled. Whole-blood gene expression data were obtained before biologic administration. Patients were categorized into remission (REM) and nonremission (NON-REM) groups according to CDAI at 6 months of biologic therapy. We employed Gene Set Enrichment Analysis (GSEA) to identify functional gene sets differentially expressed between these two groups for each biologic. Then, we compiled "signature scores" for these gene sets, and the prediction performances were assessed. Results: GSEA showed that inflammasome genes were significantly upregulated with IFX in the NON-REM group compared with the REM group. With TCZ in the REM group, B-cell-specifically expressed genes were upregulated. RNA elongation, apoptosis-related, and NK-cell-specifically expressed genes were upregulated with ABT in the NON-REM group. Logistic regression analyses showed that "signature scores" of inflammasomes, B-cell-specifically expressed, and NK-cell-specifically expressed genes were significant, independently predictive factors for treatment outcome with IFX, TCZ, and ABT, respectively. The AUCs of ROC curves of these signature scores were 0.637, 0.796, and 0.768 for IFX, TCZ, and ABT, respectively. Conclusions: We have identified original gene expression predictive signatures uniquely underlying the therapeutic effects of IFX, TCZ, and ABT. This is, to our knowledge, the first attempt to predict therapeutic effects of three drugs concomitantly using a unified gene expression test platform.

AB - Background: According to EULAR recommendations, biologic DMARDs (bDMARDs) such as tumor necrosis factor inhibitor, tocilizumab (TCZ), and abatacept (ABT) are in parallel when prescribing to rheumatoid arthritis (RA) patients who have shown insufficient response to conventional synthetic DMARDs. However, most prediction studies of therapeutic response to bDMARDs using gene expression profiles were focused on a single bDMARD, and consideration of the results from the perspective of RA pathophysiology was insufficient. The aim of this study was to identify the specific molecular biological features predicting the therapeutic outcomes of three bDMARDs (infliximab [IFX], TCZ, and ABT) by studying blood gene expression signatures of patients before biologic treatment in a unified test platform. Methods: RA patients who responded inadequately to methotrexate and were later commenced on any one of IFX (n = 140), TCZ (n = 38), or ABT (n = 31) as their first biologic between May 2007 and November 2011 were enrolled. Whole-blood gene expression data were obtained before biologic administration. Patients were categorized into remission (REM) and nonremission (NON-REM) groups according to CDAI at 6 months of biologic therapy. We employed Gene Set Enrichment Analysis (GSEA) to identify functional gene sets differentially expressed between these two groups for each biologic. Then, we compiled "signature scores" for these gene sets, and the prediction performances were assessed. Results: GSEA showed that inflammasome genes were significantly upregulated with IFX in the NON-REM group compared with the REM group. With TCZ in the REM group, B-cell-specifically expressed genes were upregulated. RNA elongation, apoptosis-related, and NK-cell-specifically expressed genes were upregulated with ABT in the NON-REM group. Logistic regression analyses showed that "signature scores" of inflammasomes, B-cell-specifically expressed, and NK-cell-specifically expressed genes were significant, independently predictive factors for treatment outcome with IFX, TCZ, and ABT, respectively. The AUCs of ROC curves of these signature scores were 0.637, 0.796, and 0.768 for IFX, TCZ, and ABT, respectively. Conclusions: We have identified original gene expression predictive signatures uniquely underlying the therapeutic effects of IFX, TCZ, and ABT. This is, to our knowledge, the first attempt to predict therapeutic effects of three drugs concomitantly using a unified gene expression test platform.

KW - BDMARDs (biologic agents)

KW - Gene expression

KW - Prediction

KW - Rheumatoid arthritis

UR - http://www.scopus.com/inward/record.url?scp=84979524509&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84979524509&partnerID=8YFLogxK

U2 - 10.1186/s13075-016-1052-8

DO - 10.1186/s13075-016-1052-8

M3 - Article

C2 - 27435242

AN - SCOPUS:84979524509

VL - 18

JO - Arthritis Research and Therapy

JF - Arthritis Research and Therapy

SN - 1478-6354

IS - 1

M1 - 159

ER -