You are here

Article Text

Article menu
Basic and translational research
Extended report
Genome-wide association analysis of anti-TNF drug response in patients with rheumatoid arthritis
  1. Maša Umiċeviċ Mirkov1,
  2. Jing Cui2,
  3. Sita H Vermeulen1,3,
  4. Eli A Stahl2,
  5. Erik J M Toonen4,
  6. Remco R Makkinje1,
  7. Annette T Lee5,
  8. Tom W J Huizinga6,
  9. Renee Allaart6,
  10. Anne Barton7,8,
  11. Xavier Mariette9,
  12. Corinne Richard Miceli9,
  13. Lindsey A Criswell10,
  14. Paul P Tak11,12,
  15. Niek de Vries11,12,
  16. Saedis Saevarsdottir13,
  17. Leonid Padyukov13,
  18. S Louis Bridges14,
  19. Dirk-Jan van Schaardenburg15,16,
  20. Tim L Jansen17,
  21. Ellen A J Dutmer18,
  22. Mart A F J van de Laar19,
  23. Pilar Barrera17,
  24. Timothy R D J Radstake20,
  25. Piet L C M van Riel17,
  26. Hans Scheffer1,
  27. Barbara Franke1,21,
  28. Han G Brunner1,
  29. Robert M Plenge2,
  30. Peter K Gregersen5,
  31. Henk-Jan Guchelaar22,
  32. Marieke J H Coenen1
  1. 1Department of Human Genetics, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  2. 2Department of Rheumatology, Immunology and Allergy, Brigham and Women's Hospital and Broad Institute, Boston, Massachusetts, USA
  3. 3Department of Epidemiology, Biostatistics and HTA, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  4. 4Department of Medicine, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  5. 5Robert S Boas Center for Genomics and Human Genetics, The Feinstein Institute for Medical Research, North Shore -LIJ, Manhasset, New York, USA.
  6. 6Department of Rheumatology, Leiden University Medical Center, Leiden, The Netherlands
  7. 7Arthritis Research UK Epidemiology Unit, Manchester Academy of Health Sciences, The University of Manchester, Manchester, UK
  8. 8NIHR Manchester Musculoskeletal Biomedical Research Unit, Central Manchester University Hospitals NHS Foundation Trust, Manchester Academic Health Science Centre, Manchester, UK
  9. 9Department of Rheumatology, Université Paris-Sud, AP-HP, Hôpitaux Universitaires Paris-Sud, INSERM U1012, Le Kremlin Bicêtre, Paris, France
  10. 10Rosalind Russell Medical Research Center for Arthritis, Department of Medicine, University of California, San Francisco, California, USA
  11. 11Department of Clinical Immunology and Rheumatology, AMC/University of Amsterdam, Amsterdam, The Netherlands
  12. 12GlaxoSmithKline, Stevenage, UK
  13. 13Rheumatology Unit, Department of Medicine, Karolinska Institutet/Karolinska University Hospital, Stockholm, Sweden
  14. 14Division of Clinical Immunology and Rheumatology, University of Alabama at Birmingham, Birmingham, USA
  15. 15Department of Rheumatology, Reade Centre for Rehabilitation and Rheumatology (formerly Jan van Breemen Institute), Amsterdam, The Netherlands
  16. 16Department of Rheumatology, VU University Medical Centre, Amsterdam, The Netherlands
  17. 17Department of Rheumatology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  18. 18Department of Rheumatology, Gelderse Vallei Hospital, Ede, The Netherlands
  19. 19Department of Rheumatology and Clinical Immunology, Arthritis Center Twente, University Twente & Medisch Spectrum Twente, Enschede, The Netherlands
  20. 20Department of Rheumatology & Clinical Immunology, University Medical Center Utrecht & Wilhemina Children's Hospital, Utrecht, The Netherlands
  21. 21Department of Psychiatry, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
  22. 22Department of Clinical Pharmacy and Toxicology, Leiden University Medical Center, Leiden, The Netherlands
  1. Correspondence to Dr Marieke Coenen, Department of Human Genetics (855), Nijmegen Centre for Evidence Based Practice and Institute for Genetic and Metabolic Disease, Radboud University Nijmegen Medical Centre, PO Box 9101, Nijmegen 6500 HB, The Netherlands; m.coenen{at}gen.umcn.nl

Abstract

Background Treatment strategies blocking tumour necrosis factor (anti-TNF) have proven very successful in patients with rheumatoid arthritis (RA). However, a significant subset of patients does not respond for unknown reasons. Currently, there are no means of identifying these patients before treatment. This study was aimed at identifying genetic factors predicting anti-TNF treatment outcome in patients with RA using a genome-wide association approach.

Methods We conducted a multistage, genome-wide association study with a primary analysis of 2 557 253 single-nucleotide polymorphisms (SNPs) in 882 patients with RA receiving anti-TNF therapy included through the Dutch Rheumatoid Arthritis Monitoring (DREAM) registry and the database of Apotheekzorg. Linear regression analysis of changes in the Disease Activity Score in 28 joints after 14 weeks of treatment was performed using an additive model. Markers with p<10−3 were selected for replication in 1821 patients from three independent cohorts. Pathway analysis including all SNPs with p<10−3 was performed using Ingenuity.

Results 772 markers showed evidence of association with treatment outcome in the initial stage. Eight genetic loci showed improved p value in the overall meta-analysis compared with the first stage, three of which (rs1568885, rs1813443 and rs4411591) showed directional consistency over all four cohorts studied. We were unable to replicate markers previously reported to be associated with anti-TNF outcome. Network analysis indicated strong involvement of biological processes underlying inflammatory response and cell morphology.

Conclusions Using a multistage strategy, we have identified eight genetic loci associated with response to anti-TNF treatment. Further studies are required to validate these findings in additional patient collections.

  • Anti-TNF
  • Gene Polymorphism
  • Pharmacogenetics
  • Rheumatoid Arthritis

https://doi.org/10.1136/annrheumdis-2012-202405

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Rheumatoid arthritis (RA) is a chronic, systemic inflammatory disease characterised by polyarthritis, joint damage and functional disability.1 It cannot be cured, and treatment is directed towards reducing the symptoms associated with the disease.

    Tumour necrosis factor α (TNFα) is a pleiotropic, proinflammatory and immunoregulatory cytokine that plays a crucial role in RA.2 ,3 The introduction of TNF-blocking agents, such as infliximab, etanercept and adalimumab, revolutionised the treatment of RA, most notably because of the excellent clinical efficacy and ability of these agents to prevent further structural damage in patients who failed to respond to treatment with conventional disease-modifying antirheumatic drugs (DMARDs).4 ,5 Despite this success, a substantial proportion of patients with RA (∼30%) treated with TNF inhibitors do not display any significant clinical improvement.6 ,7 Given the expensive treatment regimen and the potential side effects associated with the treatment, the idea of a priori prediction of response to anti-TNF agents in patients with RA is a highly relevant topic.8 ,9

    Studies of clinical parameters and biomarkers have identified several factors that influence anti-TNF treatment outcome, including concurrent use of DMARDs, lower baseline Health Assessment Questionnaire score, gender, smoking, serological status, TNF levels at the site of inflammation, and the synovial microarchitecture. However, these factors explain only a relatively small proportion of the observed variance in response (R2=17–29%) and are therefore not suitable to be used as predictors in a clinical setting.9–12 In addition, effort has been put into the identification of genetic markers predicting anti-TNF treatment outcome. Most of these studies are candidate gene based, focusing on polymorphisms in genes known to be involved in RA pathogenesis and genes implicated in TNFα signalling pathways.13 The most thoroughly investigated gene is TNFA, encoding TNFα, the target of anti-TNF treatment. Initial studies suggested a role for a variant in the promoter of the gene (−308G>A) in anti-TNF response, although recent meta-analyses do not support this association.14 ,15 So far, using the candidate gene approach, the most convincing evidence of association with response to anti-TNF therapy in patients with RA is found for an RA risk allele at the PTPRC gene locus.16 ,17

    A number of additional potential candidate loci have been suggested on the basis of results from three genome-wide association studies (GWAS).18–20 In a GWAS of 566 patients with RA, Plant et al19 found evidence of association at seven genetic loci with response to TNF blockade, two of which mapped within genes: PDZ domain-containing protein 2 (PDZD2) and eyes absent homologue 4 (EYA4). In a small study (n=89) by Liu et al,18 association was reported for markers in the MAFB and PON1 gene regions as well as in a region of chromosome 9 that contains the interferon kappa (IFN-κ), MOBKL2B and C9orf72 loci. The most compelling candidate for involvement in anti-TNF treatment response in this study is IFN-κ, since type I IFNs play a definite role in inflammatory disease and autoimmunity.21 However, these results could not be replicated by others.20 ,22 Krintel et al20 reported associations of single-nucleotide polymorphisms (SNPs) within a non-coding region surrounded by the TLR4 gene and the DBC1 gene and a marker within the FOXP1 gene with treatment outcome in a cohort of 196 Danish patients.

    To determine whether the reported loci reflect true associations, and to search for novel loci that influence differential response to anti-TNF therapy, we performed a GWAS in a cohort of 882 Dutch patients with RA receiving anti-TNF therapy.

    Materials and methods

    Patients and study design

    A multistage GWAS was performed including 984 patients with RA receiving anti-TNF medication (stage 1) with subsequent follow-up of the most significant signals in two replication cohorts (stage 2 (n=954) and 3 (n=867)).

    For the initial genome-wide association analysis, patients were recruited through a collaborative effort in which 669 patients were included as part of the Dutch Rheumatoid Arthritis Monitoring (DREAM) registry (http://www.dreamregistry.nl) and 315 patients were enrolled through the database of ApotheekZorg, which facilitates the Dutch distribution of adalimumab. All patients were diagnosed with RA according to the 1987 revised American College of Rheumatology (ACR) criteria and were treated with anti-TNF according to the indications in the Netherlands: Disease Activity Score 28 (DAS28) >3.2; previous failure on at least two DMARDs, one of which has to be methotrexate; biological naïve.23 We used the DAS28 change at 3 months as outcome for our analysis. Patients who stopped treatment within the first 3 months were not included in the study. All patients gave written informed consent, and the study was approved by the ethics committees of the participating hospitals.

    For stage 2, data from 954 RA cases treated with anti-TNF were selected from nine different cohorts as part of the ACR Research and Education Foundation (REF) ‘Within Our Reach’ project: Autoimmune Biomarkers Collaborative Network (ABCoN), Academic Medical Center Amsterdam (AMC), Behandelstrategieen voor Rheumatoide Arthritis (BeSt), Biological in Rheumatoid arthritis Genetics and Genomics Study Syndicate (BRAGGSS), Brigham Rheumatoid Arthritis Sequential Study (BRASS), Epidemiological Investigation of Rheumatoid Arthritis (EIRA), Immunex Early Rheumatoid Arthritis (ERA) study, Karolinska Institutet (KI) study, READE, formerly Jan van Breemen Institute (READE) study, Treatment of Early Aggressive RA (TEAR)—this collection has been reported on previously.16 ,24

    Finally, stage 3 included two previously described cohorts: (1) Wellcome Trust Case Control Consortium (WTCCC) comprising 595 patients with RA from the UK19; (2) 272 patients with RA from France ascertained through ReAct.25

    Genotyping and preimputation quality control (QC)

    For stage 1, genotyping was performed using the Illumina HumanHap550-Duo Bead Chip or the Human660W-Quad BeadChips, according to the instructions of the manufacturer (Illumina, San Diego, California, USA).

    Preimputation QC procedures were applied using PLINK software.26 SNPs that had minor allele frequency (MAF) <0.05 and call rates <95% were excluded, as well as SNPs with extreme departures from Hardy–Weinberg equilibrium (p<1×10−5). Subsequently, QC filtering was performed at the sample level. Four samples were excluded because of gender mismatch with phenotypic data, and 21 samples because of a genotyping rate <95%. Cryptic relatedness between study participants was examined by estimating identity by descent (IBD). Seven DNA samples were excluded based on a proportion of IBD (PI-HAT)>0.125. Lastly, principal components were computed to adjust for population stratification using the EIGENSTRAT package27; 59 individuals were removed as outliers, based on the EIGENSTRAT default filter. After QC, 882 individuals were left for analysis. For the replication cohorts, the same QC criteria were used.

    Imputation

    To obtain a marker set common to all studies and to increase overall coverage of the genome, imputation was performed using HapMap2 release 21 (downloaded from http://www.sph.umich.edu/csg/abecasis/MACH/download/HapMap-r21.html). Haplotype phasing using MaCH software (http://www.sph.umich.edu/csg/abecasis/MACH/index.html)28 was followed by genotype imputation by Minimac (http://genome.sph.umich.edu/wiki/Minimac).

    Post-imputation QC criteria were MAF ≥1% and good imputation quality, which was defined as RSQR≥0.3. In total, 2 557 253 SNPs were included in the analysis.

    Stage 1 GWAS

    The additive genetic effect of each SNP allele on change in DAS28 at 3 months of treatment was estimated using linear regression analysis with adjustment for baseline DAS28 and the first three principal components derived using EIGENSTRAT. These analyses were performed using the Mach2qtl software package29 (downloaded from http://www.sph.umich.edu/csg/abecasis/MACH/download/HapMap-r21.html).

    The Dutch samples were not genotyped in one run and therefore the results were analysed using a meta-analysis approach that combines study-specific β estimates based on the fixed-effects model and using the inverse of the variance of the study-specific β estimates to weigh the contribution of each study. Calculations were performed in the METAL package (http://www.sph.umich.edu/csg/abecasis/metal). Within-study genomic control correction was applied to the variance of β estimates using λ factors specific to each study (λ1=1.013, λ2=1.016, λ3=0.996).

    SNP selection for replication in stages 2 and 3

    Markers demonstrating association with DAS28 change (p<10−3) in stage 1 were selected for replication. Pruning of hits based on linkage disequilibrium (LD) was performed before replication: all SNPs with a HapMap Utah residents with ancestry from northern and western Europe (CEU) pair-wise correlation coefficient (r2) >0.8 with the most strongly associated SNPs in a locus were eliminated. A total of 772 independent loci were left for replication. Replication analysis in stage 2 was carried out using existing GWAS scan data from the REF collection. Those SNPs that passed the p<0.05 threshold in stage 2 were further evaluated using GWAS data from two collections (WTCCC and ReAct) in stage 3. A meta-analysis using the METAL package was performed.

    Explorative analysis for functional relation between genes identified in stage 1

    All markers showing association with DAS change (p<10−3) in stage 1 were investigated for functional interactions by Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems, www.ingenuity.com) using an unsupervised analysis. IPA computes a score for each network accordingly to the fit of the user's set of input genes. The score, representing the –log (p value), indicates the likelihood of focus genes (genes harbouring associated SNPs) in a network being found together due to random chance.

    Results

    Figure 1 presents an overview of our study approach. The baseline characteristics of the patients included in the study are summarised in table 1.

    View this table:
    Table 1

    Study population characteristics

    Figure 1
    Figure 1

    Study design of a multistage genome-wide association studies (GWAS) of response to anti-tumour necrosis factor (TNF) medication in patients with rheumatoid arthritis (RA). We started out with meta-analysis of GWAS data from a Dutch cohort comprising 882 patients with RA treated with anti-TNF medication. We selected 772 single-nucleotide polymorphisms (SNPs) that reached p<0.001, which were further followed-up in stage 2 samples (N=954 individuals). Thirty-eight SNPs out of 768 investigated in stage 2 passed p<0.05 and were further investigated in two separate cohorts in stage 3 (N=595 and N=272 individuals). MAF, minor allele frequency; QC, quality control.

    Genome-wide association analysis

    In stage 1, 2 448 996 SNPs were tested for association with anti-TNF outcome in the Dutch population. Of these SNPs, 2359 showed evidence of association with treatment response (p<10−3, online supplementary table 1). LD pruning reduced the number of SNPs prioritised for replication analysis to 772.

    We aimed to replicate the findings in 954 patients from the REF collection for stage 2 of our study. A total of 768 SNPs passed the QC in the second stage replication cohort. Thirty-nine markers showed nominally significant (p<0.05) association with treatment outcome under an additive model, 20 of which demonstrated directionally consistent association and a resulting improvement in the association signal in a stage 1 and 2 combined meta-analysis (online supplementary table 2).

    In stage 3, the 39 SNP stage 2 markers were further inspected for replication in two independent GWAS, comprising 595 patients from the UK and 272 patients from France (ReAct), separately. One SNP (rs11642036) failed QC criteria in both replication cohorts, leaving 38 SNPs for analysis. None of the tested SNPs showed nominal association with treatment outcome in these cohorts (table 2). However, the meta-analysis including all cohorts showed improved association signals for eight SNPs compared with our stage 1 results, three of which, rs1568885, rs1813443 and rs4411591, demonstrate directional consistency over all four cohorts studied (table 2).

    View this table:
    Table 2

    Association results for replicated SNPs in all cohorts and final meta-analysis

    None of the SNPs previously reported to be associated with treatment outcome showed evidence of association in the Dutch stage 1 cohort (table 3).16–20

    View this table:
    Table 3

    Initial genome-wide association studies association results for loci previously reported to be associated with anti-tumour necrosis factor treatment response

    Explorative pathway analysis of stage 1

    We explored the stage 1 dataset for potential functional relationship between genes that showed evidence of association at p<10−3 with treatment outcome using the IPA. This resulted in the identification of eight networks. The highest scoring network (p=10−41) included 26 genes identified in the genome-wide association analysis stage 1, and nine additional interacting genes (figure 2). Importantly, this network is predicted to be involved in metabolic disease and biological processes underlying inflammatory response and cell morphology and contains genes implicated in TNF signalling (NFκB) and antibody formation (IgG).

    Figure 2
    Figure 2

    Top gene network derived from Ingenuity Pathway Analysis. Genes/gene products are represented graphically as nodes, and the biological relationship between two nodes is represented as an edge (line). Grey colour of the node indicates genes that were identified in the stage 1 genome-wide association studies (p<0.001), and white indicates that the molecule was added from the Ingenuity Knowledge Base. Dashed lines indicate indirect interactions; solid lines indicate direct interactions. The style of the arrows indicates specific molecular relationships (A acts on B, A binds to B), and dotted lines indicate an indirect interaction. All edges are supported by at least one reference from the literature or from canonical information stored in the Ingenuity Knowledge Base. Nodes are displayed using various shapes that represent the functional classes of the gene product (square, cytokines; diamond, enzyme; circle in a circle, complex/group; trapezium, transporter; ellipse lying, transcription regulator; ellipse standing, transmembrane receptor; dotted rectangle, ion channel; rectangle, ligand-dependent nuclear receptor; triangle, kinase; dotted square, growth factor; circle, other).

    Discussion

    In this report, we described the results of the largest GWAS of response to anti-TNF treatment in patients with RA conducted to date.

    Using a multistage study design, we identified eight genetic loci showing suggestive evidence of association (improved p values) with treatment outcome in our overall meta-analysis, with three markers (rs4411591, rs1813443 and rs1568885) showing directional consistency over all four cohorts studied. In the combined cohort, eight identified loci together explain 3.8% of the variance in the treatment response. Although no single SNP reached a genome-wide level of significance (p<5×10−8), these variants represent excellent candidates for further investigation.

    Of the eight markers with suggestive evidence of association, two map to an intergenic region in which the nearest gene is interesting in terms of its biological function. The SNPs, rs12142623 and rs4651370, are located ∼400 kb downstream from the phospholipase A2, group IVA (PLA2G4A) gene. The protein encoded by PLA2G4A is a phospholipase enzyme involved in generation of eicosanoids, molecules with regulative function in inflammatory responses. TNFα is one of the first known stimuli for PLA2G4A activation, through the action of both TNF receptor subtypes.30 It might be possible that the identified SNPs influence long-range regulatory elements.

    In addition, three of the eight identified SNPs map within genes. The marker rs4411591 maps to the Loc100130480, encoding a hypothetical protein, while rs2378945 is located in the nucleotide-binding protein-like (NUBPL) gene. NUBPL encodes a protein required for the assembly of the respiratory chain NADH dehydrogenase (complex I). Finally, rs1813443 is located in the intronic region of contactin 5 (CNTN5), a member of the immunoglobulin superfamily, which is thought to have a role in the formation of axon connections in the developing nervous system.31 Little is known about the possible involvement of these genes in inflammatory disease, and there are no apparent functional links to anti-TNF treatment outcome, yet. If association with anti-TNF response can be confirmed in additional replication studies, future functional studies are needed to prove the biological link with anti-TNF response.

    We did not find any evidence for association for the loci identified from previously published GWAS on anti-TNF response18–20 or for the PTPRC gene (table 3)16 ,17 in the Dutch patients included in our stage 1 GWAS, suggesting that these genes do not play a major role in anti-TNF treatment outcome in our population. However, the power of our study to detect an association with PTPRC loci at p<0.0005 in a set size of 882 patients was 26%.

    However, the top network constructed with IPA indicated involvement of the genes showing suggestive association in stage 1 GWAS in the following processes: cell morphology, metabolic disease and inflammatory response. This is in line with the expected biological function of anti-TNF; TNFα is an important mediator of insulin resistance and it also impedes insulin-mediated glucose uptake.32 More importantly, other studies showed a positive long-term effect of TNF antagonists on insulin resistance, which correlated with improvement in disease activity.33 ,34 The identified network also harbours two interesting interacting molecules: NFκB and IgG. NFκB is a transcription regulator that is preferentially activated by TNF and is the main downstream target of the TNF signalling pathway.35 The finding that the genes identified through the GWAS interact with IgG is particularly interesting, since response failure and side effects of anti-TNF due to immunogenicity are not rare and it has been found that infliximab antibodies are exclusively of IgG isotype.36 Both molecules might have a central role in determining the outcome of treatment with anti-TNF agents. Besides this network, we could map the most prominent associations from stage 1 to the VAV1 and SPRED2 genes. VAV1 has been found to protect T cells from Fas-mediated apoptosis in Jurkat leukaemia T cells,37 and it has been confirmed that patients with RA show differential sensitivity to apoptosis of peripheral blood lymphocytes induced by anti-TNF therapy.38 SPRED2 is a known RA risk locus.39 This network might represent new leads and new additional candidates for future research.

    Our study, which included 2703 patients with RA treated with anti-TNF agents, is the largest GWAS of treatment outcome to date. However, our sample size still remains modest compared with genetic studies of risk of RA40 ,41 and other complex traits.42 ,43 Also, there are important aspects that may affect our results. There is considerable disease heterogeneity in RA. In our dataset, there is a difference in, for example, the number of women included in the studies, but also in type of anti-TNF treatment and co-medication use (table 1). Furthermore, the REF collection used for replication in stage 2 consists of nine smaller cohorts from several populations. Hence, combining results across four different cohorts of patients that are rather diverse in subject ascertainment and assessment and previous treatment can lead to different effect estimates among studies and false negative results. In addition, the DAS28 score, used as the measure of treatment outcome in our study, is a composite score including four measures: swollen and tender joint counts, erythrocyte sedimentation rate, and self-reported general health. This score is a powerful tool for measuring treatment response in a clinical setting. However, it is likely that this complex measure is influenced by genetic effects that are individually modest and would require large sample sizes to be detected.

    In the present study, we have identified eight genetic loci that show evidence of influencing anti-TNF treatment response based on a multistage approach in a population of 2703 Caucasian patients with RA. Our findings require further validation in independent cohorts and/or at a functional level.

    Acknowledgments

    We are indebted to Professor Lars Klareskog and Professor Lars Alfredsson for providing samples from the EIRA study, as well as Dr Johan Askling, Sara Wedrén and the Swedish Rheumatology Register, for their help with the follow-up data. We thank Alejandro Arias-Vasquez for help in the design of the analysis.

    References

    1. ↵
      1. McInnes IB,
      2. Schett G
      . Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol 2007;7:429–42.
      OpenUrlCrossRefPubMed
    2. ↵
      1. Tracey D,
      2. Klareskog L,
      3. Sasso EH,
      4. et al
      . Tumor necrosis factor antagonist mechanisms of action: a comprehensive review. Pharmacol Ther 2008;117:244–79.
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Feldmann M,
      2. Maini SR
      . Role of cytokines in rheumatoid arthritis: an education in pathophysiology and therapeutics. Immunol Rev 2008;223:7–19.
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Maini R,
      2. St Clair EW,
      3. Breedveld F,
      4. et al
      . Infliximab (chimeric anti-tumour necrosis factor alpha monoclonal antibody) versus placebo in rheumatoid arthritis patients receiving concomitant methotrexate: a randomised phase III trial. ATTRACT Study Group. Lancet 1999;354:1932–9.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Keystone EC,
      2. Kavanaugh AF,
      3. Sharp JT,
      4. et al
      . Radiographic, clinical, and functional outcomes of treatment with adalimumab (a human anti-tumor necrosis factor monoclonal antibody) in patients with active rheumatoid arthritis receiving concomitant methotrexate therapy: a randomized, placebo-controlled, 52-week trial. Arthritis Rheum 2004;50:1400–11.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Klareskog L,
      2. van der Heijde D,
      3. de Jager JP,
      4. et al
      . Therapeutic effect of the combination of etanercept and methotrexate compared with each treatment alone in patients with rheumatoid arthritis: double-blind randomised controlled trial. Lancet 2004;363:675–81.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Maini RN,
      2. Breedveld FC,
      3. Kalden JR,
      4. et al
      . Therapeutic efficacy of multiple intravenous infusions of anti-tumor necrosis factor alpha monoclonal antibody combined with low-dose weekly methotrexate in rheumatoid arthritis. Arthritis Rheum 1998;41:1552–63.
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      1. Tak PP
      . A personalized medicine approach to biologic treatment of rheumatoid arthritis: a preliminary treatment algorithm. Rheumatology (Oxford) 2012;51:600–9.
      OpenUrlAbstract/FREE Full Text
    9. ↵
      1. Hyrich KL,
      2. Watson KD,
      3. Silman AJ,
      4. et al
      . Predictors of response to anti-TNF-alpha therapy among patients with rheumatoid arthritis: results from the British Society for Rheumatology Biologics Register. Rheumatology (Oxford) 2006;45:1558–65.
      OpenUrlAbstract/FREE Full Text
    10. ↵
      1. Atzeni F,
      2. Antivalle M,
      3. Pallavicini FB,
      4. et al
      . Predicting response to anti-TNF treatment in rheumatoid arthritis patients. Autoimmun Rev 2009;8:431–7.
      OpenUrlCrossRefPubMed
    11. ↵
      1. Potter C,
      2. Hyrich KL,
      3. Tracey A,
      4. et al
      . Association of rheumatoid factor and anti-cyclic citrullinated peptide positivity, but not carriage of shared epitope or PTPN22 susceptibility variants, with anti-tumour necrosis factor response in rheumatoid arthritis. Ann Rheum Dis 2009;68:69–74.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Klaasen R,
      2. Thurlings RM,
      3. Wijbrandts CA,
      4. et al
      . The relationship between synovial lymphocyte aggregates and the clinical response to infliximab in rheumatoid arthritis: a prospective study. Arthritis Rheum 2009;60:3217–24.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Coenen MJ,
      2. Toonen EJ,
      3. Scheffer H,
      4. et al
      . Pharmacogenetics of anti-TNF treatment in patients with rheumatoid arthritis. Pharmacogenomics 2007;8:761–73.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Pavy S,
      2. Toonen EJ,
      3. Miceli-Richard C,
      4. et al
      . Tumour necrosis factor alpha -308G->A polymorphism is not associated with response to TNFalpha blockers in Caucasian patients with rheumatoid arthritis: systematic review and meta-analysis. Ann Rheum Dis 2010;69:1022–8.
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Lee YH,
      2. Ji JD,
      3. Bae SC,
      4. et al
      . Associations between tumor necrosis factor-alpha (TNF-alpha) -308 and -238 G/A polymorphisms and shared epitope status and responsiveness to TNF-alpha blockers in rheumatoid arthritis: a metaanalysis update. J Rheumatol 2010;37:740–6.
      OpenUrlAbstract/FREE Full Text
    16. ↵
      1. Cui J,
      2. Saevarsdottir S,
      3. Thomson B,
      4. et al
      . Rheumatoid arthritis risk allele PTPRC is also associated with response to anti-tumor necrosis factor alpha therapy. Arthritis Rheum 2010;62:1849–61.
      OpenUrlPubMedWeb of Science
    17. ↵
      1. Plant D,
      2. Prajapati R,
      3. Hyrich KL,
      4. et al
      . Replication of association of the PTPRC gene with response to anti-tumor necrosis factor therapy in a large UK cohort. Arthritis Rheum 2012;64:665–70.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Liu C,
      2. Batliwalla F,
      3. Li W,
      4. et al
      . Genome-wide association scan identifies candidate polymorphisms associated with differential response to anti-TNF treatment in rheumatoid arthritis. Mol Med 2008;14:575–81.
      OpenUrlPubMedWeb of Science
    19. ↵
      1. Plant D,
      2. Bowes J,
      3. Potter C,
      4. et al
      . Genome-wide association study of genetic predictors of anti-tumor necrosis factor treatment efficacy in rheumatoid arthritis identifies associations with polymorphisms at seven loci. Arthritis Rheum 2011;63:645–53.
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Krintel SB,
      2. Palermo G,
      3. Johansen JS,
      4. et al
      . Investigation of single nucleotide polymorphisms and biological pathways associated with response to TNFalpha inhibitors in patients with rheumatoid arthritis. Pharmacogenet Genomics 2012;22:577–89.
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Selmi C,
      2. Lleo A,
      3. Zuin M,
      4. et al
      . Interferon alpha and its contribution to autoimmunity. Curr Opin Investig Drugs 2006;7:451–6.
      OpenUrlPubMedWeb of Science
    22. ↵
      1. Suarez-Gestal M,
      2. Perez-Pampin E,
      3. Calaza M,
      4. et al
      . Lack of replication of genetic predictors for the rheumatoid arthritis response to anti-TNF treatments: a prospective case-only study. Arthritis Res Ther 2010;12:R72.
      OpenUrlCrossRefPubMed
    23. ↵
      1. Kievit W,
      2. Fransen J,
      3. Adang EM,
      4. et al
      . Long-term effectiveness and safety of TNF-blocking agents in daily clinical practice: results from the Dutch Rheumatoid Arthritis Monitoring register. Rheumatology (Oxford) 2011;50:196–203.
      OpenUrlAbstract/FREE Full Text
    24. ↵
      1. Moreland LW,
      2. O'Dell JR,
      3. Paulus HE,
      4. et al
      . A randomized comparative effectiveness study of oral triple therapy versus etanercept plus methotrexate in early, aggressive rheumatoid arthritis. Arthritis Rheum 2012;64:2824–35.
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Bombardieri S,
      2. Ruiz AA,
      3. Fardellone P,
      4. et al
      . Effectiveness of adalimumab for rheumatoid arthritis in patients with a history of TNF-antagonist therapy in clinical practice. Rheumatology (Oxford) 2007;46:1191–9.
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Purcell S,
      2. Neale B,
      3. Todd-Brown K,
      4. et al
      . PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81:559–75.
      OpenUrlCrossRefPubMed
    27. ↵
      1. Price AL,
      2. Patterson NJ,
      3. Plenge RM,
      4. et al
      . Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 2006;38:904–9.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Li Y,
      2. Willer C,
      3. Sanna S,
      4. et al
      . Genotype imputation. Annu Rev Genomics Hum Genet 2009;10:387–406.
      OpenUrlCrossRefPubMedWeb of Science
    29. ↵
      1. Chen WM,
      2. Abecasis GR
      . Family-based association tests for genomewide association scans. Am J Hum Genet 2007;81:913–26.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Jupp OJ,
      2. Vandenabeele P,
      3. MacEwan DJ
      . Distinct regulation of cytosolic phospholipase A2 phosphorylation, translocation, proteolysis and activation by tumour necrosis factor-receptor subtypes. Biochem J 2003;374:453–61.
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Ogawa J,
      2. Kaneko H,
      3. Masuda T,
      4. et al
      . Novel neural adhesion molecules in the Contactin/F3 subgroup of the immunoglobulin superfamily: isolation and characterization of cDNAs from rat brain. Neurosci Lett 1996;218:173–6.
      OpenUrlCrossRefPubMed
    32. ↵
      1. Peraldi P,
      2. Hotamisligil GS,
      3. Buurman WA,
      4. et al
      . Tumor necrosis factor (TNF)-alpha inhibits insulin signaling through stimulation of the p55 TNF receptor and activation of sphingomyelinase. J Biol Chem 1996;271:13018–22.
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Huvers FC,
      2. Popa C,
      3. Netea MG,
      4. et al
      . Improved insulin sensitivity by anti-TNFalpha antibody treatment in patients with rheumatic diseases. Ann Rheum Dis 2007;66:558–9.
      OpenUrlFREE Full Text
    34. ↵
      1. Tam LS,
      2. Tomlinson B,
      3. Chu TT,
      4. et al
      . Impact of TNF inhibition on insulin resistance and lipids levels in patients with rheumatoid arthritis. Clin Rheumatol 2007;26:1495–8.
      OpenUrlCrossRefPubMedWeb of Science
    35. ↵
      1. Nagar M,
      2. Jacob-Hirsch J,
      3. Vernitsky H,
      4. et al
      . TNF activates a NF-kappaB-regulated cellular program in human CD45RA- regulatory T cells that modulates their suppressive function. J Immunol 2010;184:3570–81.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. Svenson M,
      2. Geborek P,
      3. Saxne T,
      4. et al
      . Monitoring patients treated with anti-TNF-alpha biopharmaceuticals: assessing serum infliximab and anti-infliximab antibodies. Rheumatology (Oxford) 2007;46:1828–34.
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. Yin J,
      2. Wan YJ,
      3. Li SY,
      4. et al
      . The distinct role of guanine nucleotide exchange factor Vav1 in Bcl-2 transcription and apoptosis inhibition in Jurkat leukemia T cells. Acta Pharmacol Sin 2011;32:99–107.
      OpenUrlCrossRefPubMedWeb of Science
    38. ↵
      1. Coury F,
      2. Ferraro-Peyret C,
      3. Le Cam S,
      4. et al
      . Peripheral blood lymphocytes from patients with rheumatoid arthritis are differentially sensitive to apoptosis induced by anti-tumour necrosis factor-alpha therapy. Clin Exp Rheumatol 2008;26:234–9.
      OpenUrlPubMedWeb of Science
    39. ↵
      1. Baum L,
      2. Ng HK,
      3. Wong KS,
      4. et al
      . Associations of apolipoprotein E exon 4 and lipoprotein lipase S447X polymorphisms with acute ischemic stroke and myocardial infarction. Clin Chem Lab Med 2006;44:274–81.
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Okada Y,
      2. Terao C,
      3. Ikari K,
      4. et al
      . Meta-analysis identifies nine new loci associated with rheumatoid arthritis in the Japanese population. Nat Genet 2012;44:511–6.
      OpenUrlCrossRefPubMed
    41. ↵
      1. Stahl EA,
      2. Raychaudhuri S,
      3. Remmers EF,
      4. et al
      . Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis risk loci. Nat Genet 2010;42:508–14.
      OpenUrlCrossRefPubMedWeb of Science
    42. ↵
      1. Lango Allen H,
      2. Estrada K,
      3. Lettre G,
      4. et al
      . Hundreds of variants clustered in genomic loci and biological pathways affect human height. Nature 2010;467:832–8.
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      1. Chasman DI,
      2. Pare G,
      3. Zee RY,
      4. et al
      . Genetic loci associated with plasma concentration of low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglycerides, apolipoprotein A1, and Apolipoprotein B among 6382 white women in genome-wide analysis with replication. Circ Cardiovasc Genet 2008;1:21–30.
      OpenUrlAbstract/FREE Full Text
    View Abstract

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

      Files in this Data Supplement:

    Footnotes

    • Handling editor Tore K Kvien

    • RMP, PKG and H-JG contributed equally.

      TLJ, EAD, MAvdL and PLvR on behalf of the Dutch Rheumatoid Arthritis Monitoring Registry (DREAM)

    • Contributors All authors were involved in the design, analysis and interpretation of data. All authors revised the manuscript and gave final approval for its submission.

    • Funding MJHC is supported by a grant from the Netherlands Genomics Initiative (93511014). RMP is supported by grants from the NIH (R01-AR057108, R01-AR056768, U01-GM092691, R01-AR059648), and holds a Career Award for Medical Scientists from the Burroughs Wellcome Fund. Funding for this project was provided by the American College of Rheumatology Research and Education Foundation. NdV was sponsored by CTMM, the Center for Translational Molecular Medicine (http://www.ctmm.nl), and the Dutch Arthritis Foundation, project TRACER (grant 04I-202).

    • Competing interests None.

    • Ethics approval The ethics committee of the Radboud University Nijmegen Medical Centre (Commissie Mensgebonden Onderzoek (CMO) Regio Arnhem Nijmegen) approved the study (CMO number 2004/014).

    • Patient consent Obtained.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Data sharing statement Additional unpublished data can be obtained from the corresponding author upon request.