ARTICLE TITLE

Precision Medicine in Rheumatoid Arthritis

AUTHOR NAMES AND DEGREESDr James Bluett, MBBS PhD1 +44(0)161 275 1614, james.bluett@manchester.ac.uk

Professor Anne Barton, FRCP PhD1,2 +44 (0)161 275 1638, anne.barton@manchester.ac.uk

AUTHOR AFFILIATIONS1Arthritis Research UK Centre for Genetics and Genomics, Centre for Musculoskeletal Research, The University of Manchester, United Kingdom

2NIHR Manchester Musculoskeletal Biomedical Research Unit, Central Manchester NHS Foundation Trust, Manchester Academic Health Science Centre, United Kingdom

AUTHOR CONTACT INFORMATIONArthritis Research UK Centre for Genetics and GenomicsDivision of Musculoskeletal and Dermal SciencesThe University of ManchesterManchester Academic health Science CentreRoom 2.607, Stopford BuildingOxford RoadManchesterM13 9PT

CORRESPONDING AUTHOR

Dr James Bluett

DISCLOSURE STATEMENTThe Authors have nothing to disclose.

KEY WORDS

Methotrexate, anti-TNF, anti-TNF response, genetic, genomic, rheumatoid arthritis, pharmacogenomics

KEY POINTS Treatment of RA has improved in recent years but response is not universal Clinical predictors of response alone are not sufficiently predictive to aid treatment decisions Understanding the pharmacogenomics of rheumatoid arthritis would allow more personalized

healthcare

SYNOPSIS

Treatment of Rheumatoid arthritis (RA) has substantially improved in recent years due to the development of

novel drugs. However, response is not universal for any of the treatment options and selection of an effective

therapy is currently based on a trial and error approach. Delayed treatment response increases the risk of

progressive joint damage and resultant disability and also has a significant impact on quality of life for patients.

1

The challenge for the future of healthcare is to identify which patients are most likely to benefit from which

drug early in their disease and identifying biomarkers that correlate with therapy response is, therefore, a

research priority. For many drugs, the patient’s genetic background influences response to therapy and

understanding the genetics of response to therapy in RA may allow for targeted personalised healthcare.

2

Introduction

Rheumatoid Arthritis (RA) is a heterogenous disease and can range from a mild, self-limiting arthritis to rapidly

progressive joint damage. Treatment is based on controlling inflammation and early effective therapy reduces

disability, joint damage and mortality1. A range of treatment options are available but none are universally

effective, so treatment selection is based on a ‘trial and error’ approach, trying different therapies until a drug

that induces low disease activity or remission is identified2. Time on multiple ineffective medications affects

the patient’s quality of life, may lead to irreversible joint damage3, exposes the patient to potential adverse

events and is a waste of healthcare resource. Therefore, considerable research effort has been applied to

identifying predictors of drug response to allow more rational prescribing of the drug most likely to be

effective in individual patients an approach known as precision (or stratified) medicine.

Methotrexate (MTX), is the first-line therapy for RA2, while biologic therapies target specific molecular

pathways, including the tumor necrosis factor (TNF), interleukin-6, B cell and T-cell co-stimulation pathways.

The biologic drugs are typically reserved for those with an inadequate response to non-biologic disease

modifying anti-rheumatic drugs (DMARDs)2 but there is currently no guidance on which biologic agent to use

first4. Each drug has a significant failure rate; for example, TNF inhibitors (TNFi) are ineffective in up to 30%

patients5 yet remain the most commonly prescribed first-line biologic. As most research has investigated

biomarkers predictive of response to MTX and TNFi biologics, the current review will limit the focus to these

drug classes.

Treatment response is likely to be multi-factorial and influenced by clinical, psychological and biological

factors. For example, robust clinical predictors of TNFi response include disease severity, smoking status,

concomitant MTX, and patient disability, but account for a small proportion (r2=0.17) of the variance in

response and so, alone, are not useful in informing therapy selection decisions 6. There is, therefore, a need for

accurate predictors (biomarkers) of response to RA therapies to enable precision medicine - defined by

National Academy of Sciences as: the use of genomic, epigenomic, exposure and other data to define

individual patterns of disease, potentially leading to better individual treatment7.

The use of genomic variants as predictors of response has several theoretical advantages. Genetic variants are

stable and will not change due to the environment, unlike epigenetics or expression profiling. Genetic variants

3

that are associated with response are likely to be involved in key molecular pathways and can therefore

provide insight into the mechanisms of non-response. Whole genome genotyping is now economically viable

and the assays are standardized, enabling their use in the clinical setting. Indeed, genetic biomarkers are

already being used to personalize healthcare. In cystic fibrosis, for example, ivacaftor, a drug that targets the

CFTR molecule, is recommended in the 4% of patients with the G551D mutation 8 whilet in rheumatology,

screening for the enzyme TPMT, responsible for the metabolism of 6-mercaptopurine and related compounds,

is recommended to identify the 13% of the population with reduced activity and who are at increased risk of

toxicity to azathioprine9 . There are currently over 200 examples of FDA approved drugs that contain

information on genomic biomarkers that may be used to inform treatment decisions 10. While many of these

are not commonly used in clinical practice, TIPMT screening is frequently in the UK.

Studies Investigating Genomic Predictors of MTX

Given that MTX remains the treatment of choice for patients with newly-diagnosed RA, a number of studies

have investigated genes involved in the key molecular pathways affecting MTX absorption, metabolism or its

target enzymes as predictive biomarkers of response (Error: Reference source not found

The most consistent evidence for association is for the solute carrier family 19 member 1, (SLC19A1) gene, one

of several transport carriers that allow MTX to enter cells. Studies have reported that the rs1051266 variant

associates with intracellular MTX-polyglutamate levels and a recent meta-analysis of 12 studies (n=2,049)

reported an association with MTX treatment response (OR=1.49 of AA genotype, p=0.001) 11,12. Methylene

tetrahydrofolate reductase (MTHFR) is another key enzyme in the MTX pathway, and has also been extensively

investigated with a number of studies reporting associations with efficacy and toxicity. However, a meta-

analysis including 17 previous studies, revealed no association with either outcome13 and this finding has been

replicated in two subsequent meta-analyses14,15. MTX is thought to exert an anti-inflammatory effect through

inhibiting aminoimidazolecarboxamidotibonucleotide (AICAR) transformlyase (ATIC) leading to an increase in

AICAR levels and the anti-inflammatory agent adenosine16. A number of studies have associated the (ATIC 347

C>G SNP rs2372536) with toxicity17-19 but this finding has not been consistently replicated20-22.

4

As well as investigating MTX pathway genes, the major RA susceptibility gene, HLA-DRB1, has also been

studied. As the gene is associated with more severe disease23, it was hypothesized that carriers of the risk

allele would be less likely to respond to MTX monotherapy. In a study of 309 patients from an early

inflammatory polyarthritis inception cohort, presence of the HLA-DRB1 allele was associated with MTX

monotherapy inefficacy at two years (OR=3.04, p=0.02) but this finding requires replication in other data sets24.

Studies Investigating Genomic Predictors of Response to TNFi

Early candidate gene studies investigating the pharmacogenomics of TNFi therapy revealed inconsistent

findings, none of which have been robustly replicated. This review will focus on genome-wide association

studies (GWAS); candidate gene studies where findings have been replicated by at least one group and

candidate gene studies performed in sample sizes exceeding 1500 individuals (Table 1).

Whole-genome Studies

To date, five GWAS have been undertaken with the first including just 89 patients 25. 16 SNPs showed

suggestive association (p < 5 x 10-5) but none exceeded genome wide significance thresholds and none have

been replicated in subsequent, larger studies26.

5

Study n Study Design Platform SNPs for analysis

Results Validation study n SNPs Validated

Liu et al.25 89 GWAS Illumina Beadstation and Hap300 chips

283,348 16 SNPs of suggestive association

Suarez-Gestal et al.26

Krintel et al.27151196

NoneNon

Plant et al.28 566 GWAS Affymetrix GeneChip 500K

459,446 7 SNPs of suggestive association

Krintel et al.27 196 None

Krintel et al.27 196 GWAS Illumina HumanHap550K duo array

561,466 10 SNPs of suggestive association

Acosta et al.29 315 PDEA3A- SLC01C1 (OR=2.63, p=1.74x10-5)

Mirkov et al.30 882 GWAS HumanHap550-Duo /Human660W-Quad BeadChips

2,557,253 No SNP of suggestive assocation

Cui et al.31 2,706 GWAS Various > 2,000,000 1 SNP of suggestive association in etanercept-treated cohort

Cui et al.31 139 None

Cui et al.32 1,283 Candidate gene study

Various 31 1 SNP of suggestive association

Plant et al. 33

Ferreiro-Iglesias et al.34

Pappas et al.35

Zervou et al.36

1,115755233183

PTPRC (β=0.19, p=0.04)PTPRC (β=0.33, p=0.006)Not validatedNot validated

Table 1. Summary of pharmacogenomic studies investigating response to TNFi reported to date. GWAS: Genome wide association study.

6

A second GWAS undertaken by Plant et al. in 201128 included a three stage design with an initial GWAS

investigating change in DAS-28 over six months (n=566); variants with p < 10 -3 were subsequently genotyped in

an independent cohort with a subsequent meta-analysis. In stage three, variants whereby the signal was

strengthened were investigated in a third independent cohort and finally a second meta-analysis of the data

was performed. The results demonstrated seven loci associated with response but three SNPs showed an

opposite effect in the meta-analysis compared to the first stage and no SNP reached genome wide significance

(p < 5 x 10-8). Neither the Liu or Plant et al. results have been replicated subsequently27.

In 2012 Krintel et al. performed a GWAS of 196 Danish RA patients treated with TNFi, the majority of whom

were treated with infliximab and performed a subsequent meta-analysis with the Liu et al. and Plant et al.

datasets27. Response was defined as the change in Disease Activity Score on 28 joints (DAS-28) over 14 weeks.

Suggestive association was detected at the PDE3A-SLC01C1 locus, where a C>T polymorphism at rs3794271

was associated with reduced efficacy according to the EULAR criteria (OR=3.2, p=3.5x10 -6)37. A Spanish study by

Acosta et al.29 tested the same variant in 315 RA patients and replicated the association (OR=2.63, p=1.74x10 -

5). The variant was associated with response to infliximab and etanercept but not adalimumab. A subsequent

meta-analysis strengthened the association (OR-2.91, p=3.34x10-10). The PDE3A gene encodes a

phosphodiesterase (PDE), inhibition of which suppresses TNF production in lipopolysaccharide stimulated

monocytes38. The association was not reported in previous GWA studies but the variant was not tested by the

Plant et al. study. However, a subsequent study in a UK population found no evidence for association39.

A multi-stage GWAS in 2013 recruited 882 Dutch patients and two further validation cohorts (n=954 and 867

respectively)30 through international collaboration. Response was defined as three month change in DAS-28, a

shorter time period than previous studies, but no variants were associated even at suggestive association

thresholds (p < 5x10-5).

In 2013 Cui et al. performed the largest GWAS to date31. Following international collaboration, 2,706 RA

patients from 13 different cohorts treated with etanercept, infliximab or adalimumab were investigated.

Response was defined as the change in DAS-28 at 3-12 months. No association reaching genome wide

significance was detected. A subset analysis revealed SNP rs6427528 nearing genome wide significance

(p=8x10-8) in the etanercept -treated group. rs6427528 is thought to disrupt transcription binding site motifs of

CD84 and is associated with higher CD84 expression in peripheral blood mononuclear cells. CD84 is involved in

7

T-cell activation and maturation40 and acts as a costimulatory molecule for IFN-γ secretion41. Despite the strong

initial association, the SNP failed to replicate in Portuguese and Japanese cohorts (n=290).

Recently, a rigorous community-based assessment of the utility of SNP data for predicting anti-TNF treatment

efficacy in RA patients was performed in the context of a DREAM Challenge

(http://www.synapse.org/RA_Challenge) 42. This approach enabled the comparative evaluation of treatment

response predictions developed by 73 research groups using the most comprehensive available data on TNFi

response and genome-wide data. Unfortunately, no significant genetic contribution to prediction accuracy

was observed.

PTPRC, HLA-DRB1 and Response to TNFi

In 2010 Cui et al. hypothesised that genetic factors associated with RA susceptibility and severity may also be

important in predicting treatment response32. The authors investigated Caucasian individuals (n=1,283)

receiving infliximab, etanercept or adalimumab. Response was defined according to the EULAR criteria at 3-12

months post treatment. An association within the PTPRC gene was reported (rs10919563; OR=0.55, p=10-5)

with the strongest effect in the seropositive cohort. Subgroup analysis revealed PTPRC was associated with

infliximab and etanercept response but not adalimumab. PTPRC is a transmembrane receptor like molecule

that regulates T and B cell antigen receptor signalling43 and is a mediator of TNF secretion from monocytes44.

The PTPRC association has been replicated by two independent studies33,34 but two other studies failed to

replicate the association35,36, possibly due to lack of power as the two negative studies had smaller sample sizes

(Table 1). The PTPRC variant accounted for only 0.5% of the variance in response to TNFi and will therefore not

be clinically useful at the individual patient level.

The shared epitope HLA-DRB1 amino acid at positions 11, 71 and 74 confer the largest susceptibility risk of RA

and are associated with disease progression45. In a candidate gene study investigating the association between

loci within the HLA DRB1 gene and EULAR response in 1,846 RA patients treated with TNFi, a VKA haplotype at

the three amino acid positions was also significantly associated with improved EULAR response (OR=1.23,

p=0.007)23. The work suggests that the same variants may be associated with susceptibility, severity and

treatment response but with a sequential reduction in effect size, meaning that larger sample sizes are

required for treatment response studies.

8

Discussion

Despite the large number of studies investigating the pharmacogenetics and genomics of RA, results to date

have been disappointing and have not yielded a change in clinical practice. There are a number of possible

reasons for this. It could be argued that there is no genetic heritability for treatment response in RA and the

results to date are false-positives, but two studies have reported that there is detectable heritability 46,47. This,

therefore, begs the question of where the missing heritability is (see article by Laufer et al.) and there are a

number of explanations for the failure to consistently detect genetic predictors of response.

First, the majority of studies have investigated common variants, with limited power to detect the effect of

rare variants on treatment response. Technology is advancing at a staggering pace and with the reducing cost

of whole genome sequencing, it is now economically viable to evaluate the effect of rare variants however, a

recent exon sequencing study of candidate genes found no evidence for association with TNFi response in

~1,000 individuals48 .

Second, RA is not a simple monogenic disease, but a polygenic disease whereby environmental and multiple

genetic loci increase the individuals risk of developing disease. It is therefore likely that treatment response is

in part due to the multiple effect of many genetic variants, each of small effect size. A recent study used

simulations to show that successful application of common polygenic modelling approaches would require

sample sizes greater than 1,000 individuals for traits with less than 50% heritability 49 yet few studies, to date,

have included such numbers.

Third, response in RA is difficult to capture and a number of composite scores are used in clinical practice. The

majority of research studies also use these measures, including DAS-28, ACR or EULAR response criteria, which

include subjective measures of disease and are known to have a placebo effect50. Previous research has

reported that the swollen joint count and ESR are response markers that are associated with the greatest

genetic influence and that psychological factors correlate with the subjective visual analogue score (VAS) 47,51.

Some authors have used imaging to objectively assess disease activity (synovitis) and to re-weight the DAS28 52

score to more accurately reflect that feature as biologic drugs target that aspect of disease but routine MR

imaging remains impractical in the clinical setting.

Fourth, adherence to therapy is a potential confounder of predictive studies that is rarely accounted for; it has

been shown, for example, that up to 20% of RA patients self-report non-adherence to TNFi therapy when

9

asked and that non-adherence is associated with poorer clinical outcome53. Thus, a patient who is genetically

predisposed to respond to treatment may be classified as a non-responder because they are not taking the

drug but, to date, no study has adjusted for inadequate adherence to treatment.

Finally, RA may not be one disease, but different diseases, with different molecular mechanisms that present

as a similar, if heterogeneous, phenotype. For example, it is known that there are significant clinical and

genetic differences between anti-cyclic citrullinated peptide (CCP)positive and anti-CCP-negative patients 54.

Including all patients who fulfil classification criteria for RA may cause admixture of RA subsets, reducing the

power of a study to detect genetic association. Furthermore, while response to TNFi therapies is broadly

similar, there are differences illustrated by the fact that patients may experience inefficacy with one but

subsequently respond to another55. Differences in the mechanism of action are recognized; for example,

infliximab and adalimumab are licensed for Crohn’s disease whereas etanercept is ineffective 56, is associated

with a lower risk of development of tuberculosis57 and appears to be the least immunogenic, with a reported

incidence of 1-18% of anti-drug antibodies without an effect on response58. Despite this, the majority of studies

investigate response by grouping together all TNFis, potentially reducing the power to detect association in

subgroups.

The Future

Pharmacogenetic and genomic studies have the potential to enable precision medicine by providing

biomarkers to target the right drug to the right patients. Current guidelines offer little advice as to which

biologic therapy to offer patients first but stratifying patients according to increased probability of response to

a particular drug would be of major benefit4. However, studies to date have illustrated that it is unlikely that a

single genetic variant will be highly and confidently predictive of an individual’s response to drug therapy in

RA. Progress in this area will require larger sample sizes with well described patient cohorts to allow for

adjustment of confounding clinical factors such as non-adherence, anti-drug antibodies, disease severity and

smoking status. This would facilitate adequately powered studies to evaluate the effect of genetic variants on

different TNFi therapies and gene-gene interactions. Alternative measures of response that are able to

objectively characterise true responders are also required.

10

Patients are currently treated in a trial and error approach until an effective therapy is identified, but each

cycle of trying a drug destined to be ineffective increases the probability of adverse events, cost to the state,

patient dissatisfaction and the development of disability. It is therefore vital that further research is conducted

to develop precision medicine approaches, moving medicine into the 21st century.

References

1. Scire CA, Lunt M, Marshall T, et al. Early remission is associated with improved survival in patients with inflammatory polyarthritis: results from the Norfolk Arthritis Register. Annals of the rheumatic diseases. Sep 2014;73(9):1677-1682.

2. Smolen JS, Landewe R, Breedveld FC, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2013 update. Annals of the rheumatic diseases. Mar 2014;73(3):492-509.

3. Finckh A, Liang MH, van Herckenrode CM, et al. Long-term impact of early treatment on radiographic progression in rheumatoid arthritis: A meta-analysis. Arthritis Rheum. Dec 15 2006;55(6):864-872.

4. National Institute for Health and Care Excellence. TA375: Adalimumab, etanercept, infliximab, certolizumab pegol, golimumab, tocilizumab and abatacept for rheumatoid arthritis not previously treated with DMARDs or after conventional DMARDs only have failed. 2016. Accessed August 2016.

5. Gibbons LJ, Hyrich KL. Biologic therapy for rheumatoid arthritis: clinical efficacy and predictors of response. BioDrugs. 2009;23(2):111-124.

6. Hyrich KL, Watson KD, Silman AJ, 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, England). Dec 2006;45(12):1558-1565.

7. National Academy of Sciences. Toward Precision Medicine: Building a Knowledge Network for Biomedical Research and a New Taxonomy of Disease. Washington DC: National Academy of Sciences.; 2011.

8. North of England Specialised Commissioning Group. Clinical Commissioning Policy: Ivacaftor for Cystic Fibrosis. 2013; https://www.england.nhs.uk/wp-content/uploads/2013/04/a01-p-b.pdf. Accessed August 2016.

9. Reuther LO, Vainer B, Sonne J, et al. Thiopurine methyltransferase (TPMT) genotype distribution in azathioprine-tolerant and -intolerant patients with various disorders. The impact of TPMT genotyping in predicting toxicity. Eur J Clin Pharmacol. Jan 2004;59(11):797-801.

10. US Food and Drug Administration. Table of Pharmacogenomic Biomarkers in Drug Labeling. 2016; http://www.fda.gov/drugs/scienceresearch/researchareas/pharmacogenetics/ucm083378.htm. Accessed August 2016.

11. Li X, Hu M, Li W, et al. The association between reduced folate carrier-1 gene 80G/A polymorphism and methotrexate efficacy or methotrexate related-toxicity in rheumatoid arthritis: A meta-analysis. Int Immunopharmacol. Sep 2016;38:8-15.

12. Dervieux T, Kremer J, Lein DO, et al. Contribution of common polymorphisms in reduced folate carrier and gamma-glutamylhydrolase to methotrexate polyglutamate levels in patients with rheumatoid arthritis. Pharmacogenetics. Nov 2004;14(11):733-739.

13. Owen SA, Lunt M, Bowes J, et al. MTHFR gene polymorphisms and outcome of methotrexate treatment in patients with rheumatoid arthritis: analysis of key polymorphisms and meta-analysis of C677T and A1298C polymorphisms. The pharmacogenomics journal. Apr 2013;13(2):137-147.

11

14. Spyridopoulou KP, Dimou NL, Hamodrakas SJ, et al. Methylene tetrahydrofolate reductase gene polymorphisms and their association with methotrexate toxicity: a meta-analysis. Pharmacogenet Genomics. Feb 2012;22(2):117-133.

15. Lee YH, Song GG. Associations between the C677T and A1298C polymorphisms of MTHFR and the efficacy and toxicity of methotrexate in rheumatoid arthritis: a meta-analysis. Clin Drug Investig. 2010;30(2):101-108.

16. Cronstein BN, Naime D, Ostad E. The antiinflammatory mechanism of methotrexate. Increased adenosine release at inflamed sites diminishes leukocyte accumulation in an in vivo model of inflammation. J Clin Invest. Dec 1993;92(6):2675-2682.

17. Dervieux T, Greenstein N, Kremer J. Pharmacogenomic and metabolic biomarkers in the folate pathway and their association with methotrexate effects during dosage escalation in rheumatoid arthritis. Arthritis Rheum. Oct 2006;54(10):3095-3103.

18. Weisman MH, Furst DE, Park GS, et al. Risk genotypes in folate-dependent enzymes and their association with methotrexate-related side effects in rheumatoid arthritis. Arthritis Rheum. Feb 2006;54(2):607-612.

19. Grabar PB, Rojko S, Logar D, et al. Genetic determinants of methotrexate treatment in rheumatoid arthritis patients: a study of polymorphisms in the adenosine pathway. Annals of the rheumatic diseases. May 2010;69(5):931-932.

20. Takatori R, Takahashi KA, Tokunaga D, et al. ABCB1 C3435T polymorphism influences methotrexate sensitivity in rheumatoid arthritis patients. Clin Exp Rheumatol. Sep-Oct 2006;24(5):546-554.

21. Stamp LK, Chapman PT, O'Donnell JL, et al. Polymorphisms within the folate pathway predict folate concentrations but are not associated with disease activity in rheumatoid arthritis patients on methotrexate. Pharmacogenet Genomics. Jun 2010;20(6):367-376.

22. Owen SA, Hider SL, Martin P, et al. Genetic polymorphisms in key methotrexate pathway genes are associated with response to treatment in rheumatoid arthritis patients. The pharmacogenomics journal. Jun 2013;13(3):227-234.

23. Viatte S, Plant D, Han B, et al. Association of HLA-DRB1 haplotypes with rheumatoid arthritis severity, mortality, and treatment response. Jama. Apr 28 2015;313(16):1645-1656.

24. Hider SL, Silman AJ, Thomson W, et al. Can clinical factors at presentation be used to predict outcome of treatment with methotrexate in patients with early inflammatory polyarthritis? Annals of the rheumatic diseases. Jan 2009;68(1):57-62.

25. Liu C, Batliwalla F, Li W, et al. Genome-wide association scan identifies candidate polymorphisms associated with differential response to anti-TNF treatment in rheumatoid arthritis. Mol Med. Sep-Oct 2008;14(9-10):575-581.

26. Suarez-Gestal M, Perez-Pampin E, Calaza M, 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(2):R72.

27. Krintel SB, Palermo G, Johansen JS, et al. Investigation of single nucleotide polymorphisms and biological pathways associated with response to TNFalpha inhibitors in patients with rheumatoid arthritis. Pharmacogenet Genomics. Aug 2012;22(8):577-589.

28. Plant D, Bowes J, Potter C, 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. Mar 2011;63(3):645-653.

29. Acosta-Colman I, Palau N, Tornero J, et al. GWAS replication study confirms the association of PDE3A-SLCO1C1 with anti-TNF therapy response in rheumatoid arthritis. Pharmacogenomics. 2013;14(7):727-734.

30. Umicevic Mirkov M, Cui J, Vermeulen SH, et al. Genome-wide association analysis of anti-TNF drug response in patients with rheumatoid arthritis. Annals of the rheumatic diseases. Aug 2013;72(8):1375-1381.

12

31. Cui J, Stahl EA, Saevarsdottir S, et al. Genome-Wide Association Study and Gene Expression Analysis Identifies CD84 as a Predictor of Response to Etanercept Therapy in Rheumatoid Arthritis. PLoS Genetics. 2013;9(3).

32. Cui J, Saevarsdottir S, Thomson B, et al. Rheumatoid arthritis risk allele PTPRC is also associated with response to anti-tumor necrosis factor alpha therapy. Arthritis Rheum. Jul 2010;62(7):1849-1861.

33. Plant D, Prajapati R, Hyrich KL, et al. Replication of association of the PTPRC gene with response to anti-tumor necrosis factor therapy in a large UK cohort. Arthritis Rheum. Mar 2012;64(3):665-670.

34. Ferreiro-Iglesias A, Montes A, Perez-Pampin E, et al. Replication of PTPRC as genetic biomarker of response to TNF inhibitors in patients with rheumatoid arthritis. Pharmacogenomics J. Apr 2016;16(2):137-140.

35. Pappas DA, Oh C, Plenge RM, et al. Association of rheumatoid arthritis risk alleles with response to anti-TNF biologics: results from the CORRONA registry and meta-analysis. Inflammation. Apr 2013;36(2):279-284.

36. Zervou MI, Myrthianou E, Flouri I, et al. Lack of association of variants previously associated with anti-TNF medication response in rheumatoid arthritis patients: results from a homogeneous Greek population. PLoS ONE. 2013;8(9):e74375.

37. van Gestel AM, Prevoo ML, van 't Hof MA, et al. Development and validation of the European League Against Rheumatism response criteria for rheumatoid arthritis. Comparison with the preliminary American College of Rheumatology and the World Health Organization/International League Against Rheumatism Criteria. Arthritis Rheum. Jan 1996;39(1):34-40.

38. Prabhakar U, Lipshutz D, Bartus JO, et al. Characterization of cAMP-dependent inhibition of LPS-induced TNF alpha production by rolipram, a specific phosphodiesterase IV (PDE IV) inhibitor. Int J Immunopharmacol. Oct 1994;16(10):805-816.

39. Smith SL, Plant D, Lee XH, et al. Previously reported PDE3A-SLCO1C1 genetic variant does not correlate with anti-TNF response in a large UK rheumatoid arthritis cohort. Pharmacogenomics. May 2016;17(7):715-720.

40. Tangye SG, Nichols KE, Hare NJ, et al. Functional requirements for interactions between CD84 and Src homology 2 domain-containing proteins and their contribution to human T cell activation. J Immunol. Sep 1 2003;171(5):2485-2495.

41. Martin M, Romero X, de la Fuente MA, et al. CD84 functions as a homophilic adhesion molecule and enhances IFN-gamma secretion: adhesion is mediated by Ig-like domain 1. J Immunol. Oct 1 2001;167(7):3668-3676.

42. Sieberts SK, Zhu F, Garcia-Garcia J, et al. Crowdsourced assessment of common genetic contribution to predicting anti-TNF treatment response in rheumatoid arthritis. Nature communications. Aug 23 2016;7:12460.

43. Hermiston ML, Xu Z, Weiss A. CD45: a critical regulator of signaling thresholds in immune cells. Annu Rev Immunol. 2003;21:107-137.

44. Hayes AL, Smith C, Foxwell BM, et al. CD45-induced tumor necrosis factor alpha production in monocytes is phosphatidylinositol 3-kinase-dependent and nuclear factor-kappaB-independent. J Biol Chem. Nov 19 1999;274(47):33455-33461.

45. Plant D, Thomson W, Lunt M, et al. The role of rheumatoid arthritis genetic susceptibility markers in the prediction of erosive disease in patients with early inflammatory polyarthritis: results from the Norfolk Arthritis Register. Rheumatology (Oxford, England). Jan 2011;50(1):78-84.

46. Umicevic Mirkov M, Janss L, Vermeulen SH, et al. Estimation of heritability of different outcomes for genetic studies of TNFi response in patients with rheumatoid arthritis. Annals of the rheumatic diseases. Dec 2015;74(12):2183-2187.

13

47. Plant D, Bowes J, Orozco G, et al. Estimating heritability of response to treatment with anti-TNF biologic agents using linear mixed models. Annals of the rheumatic diseases. June 1, 2013 2013;71(Suppl 3):474-475.

48. Cui J, Diogo D, Stahl E, et al. The role of rare protein-coding variants to anti-TNF treatment response in rheumatoid arthritis Arthritis and rheumatology. 2016:In press.

49. Marshall SL, Guennel T, Kohler J, et al. Estimating heritability in pharmacogenetic studies. Pharmacogenomics. Mar 2013;14(4):369-377.

50. Strand V, Cohen S, Crawford B, et al. Patient-reported outcomes better discriminate active treatment from placebo in randomized controlled trials in rheumatoid arthritis. Rheumatology (Oxford, England). May 2004;43(5):640-647.

51. Cordingley L, Prajapati R, Plant D, et al. Impact of psychological factors on subjective disease activity assessments in patients with severe rheumatoid arthritis. Arthritis care & research. Jun 2014;66(6):861-868.

52. Baker JF, Conaghan PG, Smolen JS, et al. Development and validation of modified disease activity scores in rheumatoid arthritis: superior correlation with magnetic resonance imaging-detected synovitis and radiographic progression. Arthritis & rheumatology (Hoboken, N.J.). Apr 2014;66(4):794-802.

53. Bluett J, Morgan C, Thurston L, et al. Impact of inadequate adherence on response to subcutaneously administered anti-tumour necrosis factor drugs: results from the Biologics in Rheumatoid Arthritis Genetics and Genomics Study Syndicate cohort. Rheumatology (Oxford, England). Mar 2015;54(3):494-499.

54. van der Helm-van Mil AH, Verpoort KN, Breedveld FC, et al. Antibodies to citrullinated proteins and differences in clinical progression of rheumatoid arthritis. Arthritis Res Ther. 2005;7(5):R949-958.

55. Soliman MM, Hyrich KL, Lunt M, et al. Rituximab or a second anti-tumor necrosis factor therapy for rheumatoid arthritis patients who have failed their first anti-tumor necrosis factor therapy? Comparative analysis from the British Society for Rheumatology Biologics Register. Arthritis care & research. Aug 2012;64(8):1108-1115.

56. Sandborn WJ, Hanauer SB, Katz S, et al. Etanercept for active Crohn's disease: a randomized, double-blind, placebo-controlled trial. Gastroenterology. Nov 2001;121(5):1088-1094.

57. Dixon WG, Hyrich KL, Watson KD, et al. Drug-specific risk of tuberculosis in patients with rheumatoid arthritis treated with anti-TNF therapy: results from the British Society for Rheumatology Biologics Register (BSRBR). Annals of the rheumatic diseases. Mar 2010;69(3):522-528.

58. Emi Aikawa N, de Carvalho JF, Artur Almeida Silva C, et al. Immunogenicity of Anti-TNF-alpha agents in autoimmune diseases. Clin Rev Allergy Immunol. Apr 2010;38(2-3):82-89.

Figure 1. The MTX metabolic pathway. MTXPG: MTX polyglutamate; SLC19A1: solute carrier family 19 member 1; FH2: dihydrofolate; FH4: tetrahydrofolate; 5 -CH2-THF: 5-methyltetrahydrofolate 5,10-CH2-THF:

14

5,10-methyltetrahydrofolate; THF: tetrahydrofolate; IMP: inosine monophosphate; AMP: adenosine monophosphate; AICAR: aminoimidazolecarboxamidotibonucleotide.

15