Factors Predicting the Therapeutic Response to Methotrexate in Japanese Patients with Rheumatoid Arthritis: A Hospital-Based Cohort Study

Abstract

Methotrexate (MTX) is used widely as a first-line drug for the treatment of rheumatoid arthritis (RA) worldwide. There are large interindividual differences in the therapeutic response to MTX, but it is not known which factors influence them. We therefore investigated predictive factors associated with the therapeutic response to MTX in a hospital-based cohort study. Japanese adult RA outpatients prescribed MTX were enrolled and their characteristics were collected from the electronic medical records. The European League Against Rheumatism (EULAR) response criteria were used as the response to MTX therapy. The observation period was 1 year after beginning MTX administration. Sixteen types of single-nucleotide polymorphisms were investigated using the real-time PCR method. Associations between the MTX response and patient characteristics were evaluated using the multivariate logistic regression model. Among 70 Japanese adult RA outpatients, 52 were classified as MTX responders. In multivariate analysis, patients with the solute carrier family 19 member 1 (SLC19A1) 80G>A A/A genotype had a better response than those with the A/G or G/G genotype, and patients with the C allele of γ-glutamyl hydrolase (GGH) 16T>C had a better response than those with the T/T genotype.This study showed that the therapeutic response to MTX in Japanese RA patients was associated with the genetic polymorphisms of SLC19A1 80G>A and GGH 16T>C in actual clinical practice.

Rheumatoid arthritis (RA) is a chronic autoimmune disease characterized by systemic inflammatory status, joint destruction, disability, and pain.^1,2) Although the mechanism of RA onset is not fully understood, methotrexate (MTX) has been confirmed to reduce disease activity and delay or stabilize the development of bone erosion.^3,4) Among agents for the treatment of RA, currently MTX is the anchor drug, the most widely used disease-modifying antirheumatoid drug (DMARD), and therefore the first choice for RA treatment worldwide. However, major drawbacks are that patients show great interindividual variability in response to MTX, only about 50% show a good clinical response, and the unpredictable occurrence of side effects^5,6) forces 30% of patients to discontinue therapy.^7,8)

MTX is transported intracellularly via reduced folate carrier 1 (solute carrier family 19 member 1 (SLC19A1)).⁹⁾ Inside cells, MTX adds up to 4 additional glutamate moieties via folypolyglutamate synthetase (FPGS) and then forms MTX-polyglutamates (MTXPGs).^10–13) Subsequently, the terminal glutamate MTXPG molecules are removed via γ-glutamyl hydrolase (GGH),^14–17) returned to MTX (in the MTX monoglutamate form), and rapidly transported from the cells by multidrug-resistant proteins. Although the detailed mechanism of action of MTX remains unclear, it is considered to be a folate antagonist.¹⁸⁾ Accordingly, intracellular MTXPGs bind to dihydrofolate reductase (DHFR) and other folate pathway enzymes, thereby exerting antiinflammatory effects. Interindividual differences in MTX susceptibility are suggested to be due to genetic polymorphisms related to the capacity of transporters and enzymes that mediate the biotransformation and accumulation/elimination of MTX in the body.

Until now, there is a report that maximum serum concentration (C_max) after the first dose of weekly administration is a useful parameter for predicting the development of adverse reactions to MTX in Japanese rheumatoid arthritis patients.¹⁹⁾ Regarding factors affecting pharmacokinetics and response of MTX, it reported that food did not influence the bioavailability of low-dose orally administered methotrexate sodium tablets^20,21) and male sex was the only factor associated with a good response to MTX in around 8 mg/week of MTX that is recommended by the Japanese Ministry of Health, Labour, and Welfare in Japan at that time.²²⁾ Yamanaka et al.²³⁾ reported that extensive methotrexate use effectively suppressed RA disease activity and inhibits disability progression in a single institute-based large observational cohort (IORRA). Kudo-Tanaka et al.²⁴⁾ suggests that early therapeutic intervention with MTX could safely prevent the development of RA in patients with recent-onset undifferentiated arthritis. Hiraga et al.²⁵⁾ reported that serum MTX measurements could be useful in determining individual patient regimens because the degree of improvement in C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) correlated with the length of time the MTX concentration–time curve (AUC) remained above 0.02 µM in one week. Moreover, methotrexate-polyglutamates (MTX-PGs) in erythrocytes suggest being a potential indicator and predictor of MTX efficacy in patients with rheumatoid arthritis.²⁶⁾ In addition, Tazoe et al. reported that clinical efficacy of MTX for Japanese RA patients was associated with the expression level of a folate transporter gene.²⁷⁾ Hayashi et al. reported that patients with RA having SLC19A1 80A and GGH-401T alleles were less responsive to MTX than those with SLC19A1 80A and without GGH-401T alleles.²⁸⁾ In addition, Kato et al. reported that the genotype at 16 polymorphic sites in 11 genes (ATP-binding cassette subfamily B member 1 (ABCB1), ATP-binding cassette sub-family G member 2 (ABCG2), ATP-binding cassette sub-family C member 2 (ABCC2), SLC19A11, PCFT, SLCO1B1, methylenetetrahydrofolate reductase (MTHFR), GGH, ATIC, MTR, and MTRR) was analyzed, and the ABCB1 3435C>T, ABCB1 2677G>A/T, and MTHFR 1298A>C polymorphisms influenced the efficacy of MTX monotherapy.²⁹⁾

To date, the validated markers for predicting the MTX therapeutic response have not been established. It is necessary to develope a precise strategy for MTX treatment to overcome the large interindividual differences in its effectiveness. However, despite many efforts to identify factors predicting the response to MTX treatment, which have focused on drug disposition including single-nucleotide polymorphisms (SNPs) in genes coding for folate pathway enzymes and MTX transport into and out of cells in patients with RA, markers in relation to MTX efficacy have not been determined.^30,31)

The purpose of this study was to clarify the relationship between the therapeutic response to MTX and disease control status, identify the genetic polymorphisms for MTX-related transporters/enzymes, and obtain better supportive information to devise a strategy for precise MTX treatment in Japanese patients with RA.

METHODS

Study Design, Setting, Participants

This study used a real world clinical data in Keio University Hospital (Tokyo, Japan). We enrolled 159 patients in the base cohort, who were aged ≥18 years with a confirmed diagnosis of RA (fulfilling American College of Rheumatology [ACR] 1987 criteria³²⁾ or ACR/European League Against Rheumatism [EULAR] 2010 criteria³³⁾), who regularly visited the Division of Rheumatology, Department of Internal Medicine, Keio University Hospital, from June to September 2015, and who had a documented history of prescribed MTX and folic acid. The forms of methotrexate prescribed were Rheumatrex (Pfizer Inc., U.S.A.), Metolate (Santen Pharmaceutical Co., Ltd., Japan), or Methotrexate (Pfizer Inc.). All patients gave written informed consent for participation in this study. Patients in the base cohort who had a Disease Activity Score for 28 joints (DAS28) of ≥3.2 at the starting of MTX administration were included. We used the following expressions for calculating the DAS28:   

where T28=28 tender joints, S28=28 swollen joints, and VAS=Visual Analogue Scale. However, patients who met the following criteria were excluded: 1) use of biological DMARDs (bDMARDs) within 4 months from starting MTX administration; 2) use of conventional synthetic DMARDs (csDMARDs) except for MTX; and 3) poor treatment compliance as indicated in the electronic medical records. The observation period in this study was 1 year after the start of MTX administration or until June 30, 2016.

Genotype Determination

DNA Extraction

For genetic analysis, 10 mL of peripheral blood was collected from each RA patient in a tube containing ethylenediaminetetraacetic disodium salt and stored at −20°C until DNA extraction. DNA was extracted using the Wizard Genomic DNA Purification Kit (Promega Corporation, Madison, WI, U.S.A.). Total genomic DNA was quantified, and its purity and integrity were analyzed using Shimadzu BioSpec-nano (Shimadzu, Kyoto, Japan).

Genotyping of MTX-Related Transporter and Enzyme Genes

Genotyping of alleles of SLC19A1 80G>A (rs1051266), ABCB1 1236C>T (rs1128503), ABCB1 3435C>T (rs1045642), ABCC2 1249G>A (rs2273697), ABCC2 1058G>A (rs7080681), ABCC2 IVS23+56C>T (rs4148396), ABCG2 914C>A (rs2231142), FPGS 1994G>A (rs10106), GGH 16T>C (rs1800909), GGH 452C>T (rs11545078), GGH –401C>T (rs3758149), MTHFR 677C>T (rs1801133), MTHFR 1298A>C (rs1801131), thymidylate synthase (TYMS) 3′-untranslated region (3′-UTR) 6-bp deletion/insertion (rs16430), and ATIC 347C>G (rs2372536) was performed using the TaqMan SNP Genotyping Assay from Applied Biosystems (Foster City, CA, U.S.A.) with fluorogenic binding probes. PCR amplification with the real-time PCR method was performed in 5 µL of reaction mixture including genomic DNA 4 ng, 0.0625 µL of 40×TaqMan SNP Genotyping Assay Mix, and TaqMan Universal PCR Master Mix (Applied Biosystems) 2.5 µL. The reaction conditions for PCR were: pre-PCR for 30 s at 25°C; initial denaturation for 10 min at 95°C; 40 cycles (50 cycles for ABCB1 3435C>T and ABCG2 914C>A) at 92°C/15 s; annealing and extension for 1 min at 58°C (55°C for FPGS 1994G>A, 52°C for TYMS 3′-UTR 6-bp deletion/insertion); and post-PCR for 30 s at 25°C. The PCR system used was the StepOnePlus real-time PCR system (Applied Biosystems). The genotype of the TYMS 5′-untranslated region (5′-UTR) 28-bp tandem repeat (rs34743033) was detected using a PCR-restriction fragment length polymorphism method. Genomic DNA (100 ng) was used in 25 µL of reaction mixture with TaKaRa Taq (TaKaRa Bio, Shiga, Japan) 0.125 µL, 10× PCR buffer (Mg²⁺) 2.5 µL, deoxyribonucleotide triphosphate (dNTP) mixture (2.5 mM each) 2 µL, 1.0 µM of each primer, and glycerol 2.5 µL. The forward primer was 5′-GTG GCT CCT GCG TTT CCC CC-3′, and the reverse primer was 5′-GCT CCG AGC CGG CCA CAG GCA TGG CGC GG-3′. Amplification conditions were an initial denaturation cycle for 10 min at 95°C and 35 cycles of the following: 95°C for 30 s, 61°C for 30 s, and 72°C for 45 s; and then a final extension at 72°C for 5 min. Individuals with the 2R/2R genotype presented one fragment (210 bp), those with the 2R/3R genotype presented two fragments (210 bp and 238 bp), and those with the 3R/3R genotype presented one fragment (238 bp) when visualized on 3% agarose gels with ethidium bromide.³⁴⁾

Variables Related to Patient Characteristics

We evaluated the response to MTX using the EULAR response criteria that take into account the differences between the DAS28 at the start of MTX administration (baseline) and at 4 or 6 months. Response was defined as: good response, when the DAS28 score at the endpoint was ≤3.2 and DAS28 improvement from baseline was ≥1.2; moderate response, when DAS28 at the endpoint was between 3.2 and 5.1, with DAS28 improvement between 0.6 and 1.2; or poor response, when DAS28 at the endpoint was than >5.1 and DAS28 improvement was ≤0.6 compared with baseline. We classified patients with good and moderate responses as responders, and those with a poor response as nonresponders.

We used the following variables to investigate the factors related to response to MTX: age; gender; genetic polymorphisms; maximum MTX dosage at the endpoint; maximum folic acid dosage at the endpoint; estimated glomerular filtration rate (eGFR) at baseline; DAS28 at baseline; disease duration from onset to baseline; rheumatoid factor (RF) positivity or negativity; and anticitrullinated protein antibody (ACPA) positivity or negativity. We collected data on patient characteristics and drugs prescribed as well as laboratory data from the hospital electronic medical records. If laboratory data were lacking, we used those recorded on the nearest test day within 4 months before and after the day for which data were missing.

Statistical Analysis

Patients were classified as MTX responders or nonresponders. The distribution of categorical variables (gender, RF, ACPA, genetic polymorphisms) between groups was evaluated with χ² analysis and fisher's exact test. Between-group differences in MTX and folic acid dosage were evaluated using Student’s t-test, and age, eGFR levels at baseline, DAS28 at baseline, and disease duration were evaluated using the Mann–Whitney U-test. Tests for Hardy–Weinberg equilibrium were carried out using the χ² test. The association between response to MTX and patient characteristics (background and genetic polymorphisms of the 15 SNPs) was evaluated using the multivariate logistic regression model, calculated odds ratios (ORs) for responders, and 95% confidence intervals (CIs).

A stepwise backward-elimination procedure was used. Variables with a p value of less than 0.5 in univariate analysis were entered together into a multivariable logistic model and then removed until all retained variables had a p value of less than 0.1. Analysis was conducted including the following potential confounders: MTX dosage; folic acid dosage; eGFR at baseline; DAS28 at baseline; RF; ACPA; and genetic polymorphisms of MTX-related enzymes and transporters (SLC19A1 80G>A, ABCB1 1236C>T, ABCB1 3435C>T, ABCC2 1249G>A, ABCC2 IVS23+56C>T, ABCG2 914C>A, FPGS 1994G>A, GGH 16T>C, GGH 452C>T, GGH –401C>T, MTHFR 677C>T, MTHFR 1298A>C, TYMS 3′-UTR 6-bp deletion/insertion, TYMS 5′-UTR 28-bp tandem repeat, ATIC 347C>G). We classified all polymorphisms into 2 groups (recessive model) to investigate whether mutant alleles affected MTX response (wild-type homozygous vs. heterozygous and mutant-type homozygous; for the TYMS 5′-UTR 28-bp tandem repeat, 2R/2R+2R/3R vs. 3R/3R+3R/4R+3R/5R).

All statistical analyses were conducted using the IBM SPSS Statistics for Windows Version 23.0 J (IBM SPSS Statistics, Chicago, IL, U.S.A.).

RESULTS

Patient Cohort

Patient characteristics and the study flow diagram of patients enrolling for the efficacy judgement are shown in Table 1 and Fig. 1, respectively. A total of 70 unrelated Japanese RA patients (16 men, 64 women) meeting the selection criteria were enrolled in this study, of whom 52 were responders to MTX treatment. Data were missing on disease duration for 6 patients, DAS28 at baseline for 4, and ACPA for 2. Sixty-seven of the 70 patients were receiving folic acid. There were no statistically significant differences between responders and nonresponders in any variable except for eGFR at baseline.

Table 1. Characteristics of Patients in the Study Cohort

Characteristic	Total (n=70)	Responders (n=52)	Nonresponders (n=18)	p Value
Age (years; mean±S.D.)	58.7±12.7	59.4±12.4	57.4±14.4	0.309
Gender, female (%)	64 (91.4)	46 (88.5)	18 (100)	0.327
Disease duration (years; mean±S.D.)a)	3.33±5.97	3.32±5.44	4.14±7.93	0.644
DAS28 at baseline (mean±S.D.)b)	4.62±0.93	4.74±0.92	4.28±0.89	0.071
HAQ-DI (mean±S.D.)	0.69±0.62 (N=68)	0.63±0.61 (N=50)	0.85±0.62 (N=18)	0.198
Rheumatoid factor, positive, n (%)	47 (67.1)	36 (69.2)	11 (61.1)	0.527
Anticyclic citrullinated peptide antibody (ACPA), positivity, n (%)c)	41 (60.3)	31 (60.8)	10 (58.8)	0.866
MTX dosage (mg/week; mean±S.D.)	10.1±2.76	10.3±2.83	10±2.66	0.725
Folic acid dosage (mg/week; mean±S.D.)	6.62±4.39	6.92±4.66	6.11±4.04	0.388
eGFR at baseline (mL/min/1.73 m2; mean±S.D.)	79.3±17.5	79.4±18.9	78.5±14.6	0.009

a) Disease duration: data missing for 6 patients; b) DAS28 at baseline: data missing for 4 patients; c) ACPA: data missing for 2 patients.

Fig. 1. Flow Diagram of Patients Enrolling for the Efficacy Judgement in This Study

Genotype Frequencies at Each SNP Site Evaluated

Frequencies of each genotype are shown in Table 2. The frequencies of all genotypes except for ABCC2 1058G>A were in agreement with Hardy–Weinberg equilibrium. There were no significant differences in genotype frequencies between MTX responders and nonresponders. In our cohort, only the ABCC2 1058G>A wild-type homogenous genotype was found. The GGH 16T>C and GGH –401C>T genotypes were in strict linkage disequilibrium (r=0.977).

Table 2. Frequencies of Genotypes in MTX Responders, and Nonresponders

Variable		Total (n=70)	Responders (n=52)	Nonresponders (n=18)	p Value
SLC19A1 80G>A	G/G	16 (22.9)	10 (19.2)	6 (33.3)	0.105
	G/A	28 (40.0)	19 (36.5)	9 (50.0)
	A/A	26 (37.1)	23 (44.2)	3 (16.7)
ABCB1 1236C>T	C/C	8 (11.4)	7 (13.5)	1 (5.6)	0.427
	C/T	35 (50.0)	27 (51.9)	8 (44.4)
	T/T	27 (38.6)	18 (34.6)	9 (50.0)
ABCB1 3435C>T	C/C	23 (32.9)	18 (34.6)	5 (27.8)	0.758
	C/T	35 (50.0)	26 (50.0)	9 (50.0)
	T/T	12 (17.1)	8 (15.4)	4 (22.2)
ABCC2 1058G>A	G/G	70 (100)	52 (100)	18 (100)	—
	G/A	0 (0.0)	0 (0.0)	0 (0.0)
	A/A	0 (0.0)	0 (0.0)	0 (0.0)
ABCC2 1249G>A	G/G	53 (75.7)	39 (75)	14 (77.8)	0.700
	G/A	15 (21.4)	11 (21.2)	4 (22.2)
	A/A	2 (2.9)	2 (3.8)	0 (0.0)
ABCC2 IVS 23+56C>T	C/C	49 (70.0)	34 (65.4)	15 (83.3)	0.171
	C/T	19 (27.1)	17 (32.7)	2 (11.1)
	T/T	2 (2.9)	1 (1.9)	1 (5.6)
ABCG2 914C>A	C/C	32 (45.7)	26 (50.0)	6 (33.3)	0.470
	C/A	32 (45.7)	22 (42.3)	10 (55.6)
	A/A	6 (8.6)	4 (7.7)	2 (11.1)
FPGS 1994G>A	G/G	25 (35.7)	17 (32.7)	8 (44.4)	0.669
	G/A	36 (51.4)	28 (53.8)	8 (44.4)
	A/A	9 (12.9)	7 (13.5)	2 (11.1)
GGH 16T>C	T/T	35 (50.0)	22 (42.3)	13 (72.2)	0.082
	C/T	33 (47.1)	28 (53.8)	5 (27.8)
	C/C	2 (2.9)	2 (3.8)	0 (0.0)
GGH 452C>T	C/C	58 (82.9)	42 (80.8)	16 (88.9)	0.677
	C/T	11 (15.7)	9 (17.3)	2 (11.1)
	T/T	1 (1.4)	1 (1.9)	0 (0.0)
GGH–401C>T	C/C	36 (51.4)	23 (44.2)	13 (72.2)	0.109
	C/T	32 (45.7)	27 (51.9)	5 (27.8)
	T/T	2 (2.9)	2 (3.8)	0 (0.0)
MTHFR 677C>T	C/C	25 (35.7)	19 (36.5)	6 (33.3)	0.738
	C/T	34 (48.6)	24 (46.2)	10 (55.6)
	T/T	11 (15.7)	9 (17.3)	2 (11.1)
MTHFR 1298A>C	A/A	48 (68.6)	35 (67.3)	13 (72.2)	0.838
	A/C	19 (27.1)	15 (28.8)	4 (22.2)
	C/C	3 (4.3)	2 (3.8)	1 (5.6)
TYMS 3′-UTR 6-bp deletion/insertion	−6/−6	30 (42.9)	23 (44.2)	7 (38.9)	0.834
	−6/+6	31 (44.3)	23 (44.2)	8 (44.4)
	+6/+6	9 (12.9)	6 (11.5)	3 (16.7)
TYMS 5′-UTR 28-bp tandem repeat	2R/2R	5 (7.1)	4 (7.7)	1 (5.6)	0.359
	2R/3R	15 (21.4)	9 (17.3)	6 (33.3)
	3R/3R	50 (71.4)	39 (75)	11 (61.1)
ATIC 347C>G	C/C	43 (61.4)	32 (61.5)	11 (61.1)	0.438
	C/G	23 (32.9)	16 (30.8)	7 (38.9)
	G/G	4 (5.7)	4 (7.7)	0 (0.0)

Univariate and Multivariate Analyses

The results of univariate and multivariate analyses of factors affecting the MTX response are shown in Table 3. Data for ABCC2 1058G>A were excluded from the analysis because all patients had the ABCC2 G/G genotype. In univariate analysis of nongenetic factors, the crude OR (95% CI) of DAS28 at baseline in responders was greater than that in nonresponders (1.86 [0.94–3.69]), although the difference did not reach statistical significance. Among genetic factors, the crude OR in patients with the G allele of SLC19A1 80G>A was significantly less than in those with the A allele homozygotic genotype (0.25 [0.65–0.98]), and a significantly better therapeutic response to MTX was seen in patients with the A/A genotype in SLC19A1 80G>A. The crude ORs in patients with the wild homozygotic or wild/mutant heterozygotic genotype of ABCB1 1236C>T, ABCC2 IVS23+56C>T, FPGS 1994G>A, and GGH 452C>T compared with the mutant homozyotic genotypes were 1.89 (0.64–5.60), 2.65 (0.68–10.36), 1.65 (0.55–4.93), and 1.91 (0.38–9.66), respectively, although they did not show significant differences. Patients with the C allele of GGH 16T>C showed a significantly better therapeutic response to MTX (3.55 [1.10–11.4]).

Table 3. Results of Univariate and Multivariate Analyses

Variable		Total (n=70)	Responders (n=52)	Nonresponders (n=18)	Crude OR	Adjusted OR
Age (years)		58.7±12.7	59.4±12.4	57.4±14.4	1.01 (0.97–1.05)
RF		47 (67.1)	36 (69.2)	11 (61.1)	1.43 (0.47–4.37)
					p=0.528
ACPA, positive, n (%)a)		41 (60.3)	31 (60.8)	10 (58.8)	1.09 (0.36–3.32)
					p=0.886
Disease duration (years; mean±S.D.)b)		3.33±5.97	3.32±5.44	4.14±7.93	0.98 (0.90–1.07)
					p=0.979
DAS28 at baselinec) (mean±S.D.)		4.62±0.93	4.74±0.92	4.28±0.89	1.86 (0.94–3.69)	—
					p=0.576
MTX dosage (mg/week; mean±S.D.)		10.1±2.76	10.3±2.83	10±2.66	1.04 (0.85–1.27)
					p=0.721
Folic acid dosage (mg/week; mean±S.D.)		6.62±4.39	6.92±4.66	6.11±4.04	1.05 (0.91–1.20)
					p=0.510
eGFR at baseline (mL/min/1.73 m2; mean±S.D.)		79.3±17.5	79.4±18.9	78.5±14.6	1.00 (0.97–1.03)
					p=0.855
SLC19A1 80G>A	A/A	26 (37.1)	23 (44.2)	3 (16.7)	1.00	1.00
	G allele	44 (62.9)	29 (55.8)	15 (83.3)	0.25 (0.65–0.98)	0.15 (0.03–0.76)
					p=0.046	p=0.022
ABCB1 1236C>T	T/T	27 (38.6)	18 (34.6)	9 (50.0)	1.00
	C allele	43 (61.4)	34 (65.4)	9 (50.0)	1.89 (0.64–5.60)
					p=0.251
ABCB1 3435C>T	C/C	23 (32.9)	18 (34.6)	5 (27.8)	1.00
	T allele	47 (67.1)	34 (65.4)	13 (72.2)	0.73 (0.22–2.36)
					p=0.595
ABCC2 1249G>A	G/G	53 (75.7)	39 (75.0)	14 (77.8)	1.00
	A allele	17 (24.3)	13 (25.0)	4 (22.2)	1.17 (0.33–4.18)
					p=0.813
ABCC2 IVS 23+56C>T	C/C	49 (70.0)	34 (65.4)	15 (83.3)	1.00
	T allele	21 (30.0)	18 (34.6)	3 (16.7)	2.65 (0.68–10.36)
					p=0.162
ABCG2 914C>A	C/C	32 (45.7)	26 (50.0)	6 (33.3)	1.00
	A allele	38 (54.3)	26 (50.0)	12 (66.7)	0.50 (0.16–1.53)
					p=0.225
FPGS 1994G>A	G/G	25 (35.7)	17 (32.7)	8 (44.4)	1.00	1.00
	A allele	45 (64.3)	35 (67.3)	10 (55.6)	1.65 (0.55–4.93)	3.26 (0.84–12.7)
					p=0.372	p=0.089
GGH 16T>C	T/T	35 (50.0)	22 (42.3)	13 (72.2)	1.00	1.00
	C allele	35 (50.0)	30 (57.7)	5 (27.8)	3.55 (1.10–11.4)	4.80 (1.21–19.0)
					p=0.034	p=0.026
GGH 452C>T	C/C	58 (82.9)	42 (80.8)	16 (88.9)	1.00
	T allele	12 (17.1)	10 (19.2)	2 (11.1)	1.91 (0.38–9.66)
					p=0.437
MTHFR 677C>T	C/C	25 (35.7)	19 (36.5)	6 (33.3)	1.00
	T allele	45 (64.3)	33 (63.5)	12 (66.7)	0.87 (0.28–3.69)
					p=0.807
MTHFR 1298A>C	A/A	48 (68.6)	35 (67.3)	13 (72.2)	1.00
	C allele	22 (31.4)	17 (32.7)	5 (27.8)	1.26 (0.39–4.12)
					p=0.699
TYMS 3′-UTR 6-bp deletion/insertion	–6 bp/−6 bp	30 (42.9)	23 (44.2)	7 (38.9)	1.00
	With+6bp*d)	40 (57.1)	29 (55.8)	11 (61.1)	0.80 (0.27–2.40)
					p=0.693
TYMS 5′-UTR 28-bp tandem repeat	3R/3R	50 (71.4)	39 (75.0)	11 (61.1)	1.00	1.00
	With 2R	20 (28.6)	13 (25.0)	7 (38.9)	0.52 (0.17–1.63)	0.26 (0.06–1.10)
					p=0.265	p=0.068
ATIC 347C>G	C/C	43 (61.4)	32 (61.5)	11 (61.1)	1.00
	G allele	27 (38.6)	20 (38.5)	7 (38.9)	0.98 (0.33–2.95)
					p=0.974

a) ACPA: data missing for 2 patients; b) Disease duration: data missing for 6 patients; c) DAS28 at baseline: data missing for 4 patients; d) With +6bp: +6 bp/−6 bp and +6 bp/+6 bp.

In multivariate analysis of genetic factors, the adjusted OR in patients with the G allele of SLC19A1 80G>A was significantly less than in those with the A allele homozygotic genotype (0.15 [0.03–0.76]), and a significantly better therapeutic response to MTX was seen in patients with the A/A genotype in SLC19A1 80G>A. The adjusted OR in patients with the C allele of GGH 16T>C was significantly greater than in the TT homozygotic genotype group (4.80 [1.21–19.0]), and a significantly better therapeutic response to MTX was seen in patients with the C allele. The adjusted OR [95%CI] in patients with the A/A or G/A genotype of FPGS 1994G>A was 3.26 [0.84–12.7] compared with the G/G genotype group, and more patients with the wild-type homozygotic or wild/mutant heterozygotic genotype were MTX treatment responders, although the difference did not reach statistical significance. The adjusted OR for the TYMS 5′UTR 28-bp tandem repeat was 0.26 (0.06–1.10), and fewer patients had the 3R/3R genotype than the 2R homozygotic or 2R/3R hereozygotic genotype, although the difference was not significant. However, a significantly better response to MTX was shown by patients with the 3R/3R genotype in the TYMS 5′-UTR 28-bp tandem repeat group.

DISCUSSION

We investigated the risk factors, including patient background factors and genetic polymorphisms, associated with the therapeutic response to MTX in 70 unrelated Japanese RA patients in clinical practice. The frequencies of genetic polymorphisms in our cohort were similar to those in other reports.^28,29,35,36)

For the studies of Japanese RA patients, Tazoe et al. reported that clinical efficacy of MTX for Japanese RA patients was associated with the expression level of a folate transporter gene.²⁷⁾ Increased SLC19A11 expression may increase MTX uptake by immune cells, such as lymphocytes, and as a result, RA disease activity would be reduced.²⁷⁾ Hayashi et al. reported that patients with RA having SLC19A1 80A and GGH-401T alleles were less responsive to MTX than those with SLC19A1 80A and without GGH-401T alleles.²⁸⁾ In addition, Kato et al. reported that the genotype at 16 polymorphic sites in 11 genes (ABCB1, ABCG2, ABCC2, SLC19A1, PCFT, SLCO1B1, MTHFR, GGH, ATIC, MTR, and MTRR) was analyzed, and the ABCB1 3435C>T, ABCB1 2677G>A/T, and MTHFR 1298 A>C polymorphisms influenced the efficacy of MTX monotherapy.²⁹⁾ Until now, there is little study to clarify the relationship between the therapeutic response to MTX and disease control status, identify the genetic polymorphisms for MTX-related transporters/enzymes in Japanese RA patients using the multivariate analysis. Therefore, this is a study to obtain better supportive information to devise a strategy for precise MTX treatment in Japanese patients with RA.

In this study, only the SLC19A1 80G>A and GGH 16T>C polymorphisms were extracted as factors in multivariate analysis. For SLC19A1 80G>A, more patients had the A/A than the G/G genotype in this study, and therefore we considered the A allele as a wild-type allele and the G allele as mutant-type allele. The patients with the A/A genotype in SLC19A1 80G>A tended to have a better therapeutic response to MTX than those with the G allele homozygotic and heterozygotic genotypes. Drozdzik et al.³⁷⁾ found a better MTX response in patients with the A/A genotype of SLC19A1 80G>A. In addition, meta-analysis by Kung et al.³⁸⁾ showed that the A/A genotype group of SLC19A1 80G>A had a better response to MTX (OR 1.42 [1.04–1.93]) than patients with the G allele. Therefore, the current study supports the results of both Drozdzik et al.³⁷⁾ and Kung et al.³⁸⁾ On the other hand, from the viewpoint of the functional change in SLC19A1, it was reported that MTX uptake efficiency by cells increased in patients with the A allele.³⁹⁾

This study also found that those with the GGH 16T>C C/C and C/T genotypes had a better therapeutic response to MTX (OR 4.8 [1.21–18.0]) compared with those with the T/T genotype. van der Straaten et al.⁴⁰⁾ showed that patients with the GGH 16T>C C/C and C/T genotypes were 2.9-fold more likely to reduce the DAS by >1.2 compared with those with the T/T genotype. The present results therefore support the findings of van der Straaten et al.,⁴⁰⁾ however, it is not reported the function change by the gene variation of GGH16T>C.

No research on multivariate analysis of factors affecting the correlation between patient background characteristics including multiple genetic polymorphisms and therapeutic response using EULAR has been conducted in Japanese RA patients. Therefore, it was possible to conduct analysis excluding the influence of factors related to the therapeutic response using multivariate analysis, and this study revealed that genetic polymorphisms of SLC19A1 and GGH are also related to the response in Japanese RA patients. Takahashi et al.⁴¹⁾ reported that despite the use of similar dosages of MTX, the MTX-PG concentrations in Japanese patients were markedly higher than those observed in other studies from Europe or the U.S.A. Therefore, we believe that the effect on the therapeutic response to MTX in Japanese is greater in terms of MTX pharmacokinetics than in RA patients in Europe or the U.S.A. If so, SLC19A1 and GGH extracted as genetic variables in this study appear to affect MTX-PG concentrations as part of the pharmacokinetics of MTX. This would also support the results of Takahashi et al.⁴¹⁾

The patients with the FPGS 1994G>A A/A and G/A genotypes tended to be better therapeutic response to MTX (OR 3.26 [0.84–12.7]) than patients with the G/G genotype, although the difference was not slightly significant because of type II error. Stranzl et al.⁴²⁾ found an association between FPGS mRNA expression in peripheral blood mononuclear cells and poor response to MTX in RA patients. Yamamoto et al.⁴³⁾ reported that three polymorphisms of FPGS (rs10106, rs1054774, rs1544105) significantly influenced the MTXPG3-5/1-2 ratio in red blood cells, while polymorphisms of SLC19A1 and GGH had no impact. The order of intracellular MTXPG3-5/1-2 ratios were G/G>A/G>A/A in FPGS (rs10106). Therefore, FPGS may have a major role in regulating intracellular polyglutamation of MTX in RA patients receiving low-dose weekly administration, although the functional roles of SNPs in this gene are yet to be determined.

In addition, the 3R/3R genotype among patients with the TYMS 5′-UTR 28-bp tandem repeat was associated with a better MTX therapeutic response than the 2R allele, although the difference was not slightly significant. The same tendency was found by James et al.,⁴⁴⁾ but the opposite result was reported by Lima et al.⁴⁵⁾ For the TYMS 5′-UTR 28-bp tandem repeat, it was reported that 3R compared with 2R increased the level of mRNA expression,^46,47) and MTX dosage increases in patients with the 3R allele achieved the same therapeutic effect as in those without 3R allele.⁴⁸⁾ However, this study did not examine differences in the MTX dose required among TYMS 5′-UTR genotypes. Therefore further study regarding the association between TYMS 5′-UTR genetic polymorphisms and the clinical therapeutic effect of MTX is required.

Strengths and Limitations

Two strengths of this study were, first, that it investigated the MTX therapeutic response based on the EULAR response criteria. This was the initial study using the EULAR response plus associations with genetic factors in Japanese RA patients. Second, we extracted the factors that influence the response to MTX in daily clinical practice in Japanese RA patients. The study of the influence factor in conjunction with the efficacy of MTX by the multivariate analysis including a patient background factor and the genetic polymorphism in the Japanese RA patient had not been reported. However, this study also had several limitations, e.g., the starting date of MTX administration was unknown for several patients and DAS28 data were missing for some. We evaluated the treatment response at 6 months after beginning MTX administration using EULAR. Therefore, whether the patients were receiving a sufficient MTX dosage to show a therapeutic effect was not assessed. The most important limitation of this study, however, was the small sample size. This was a retrospective cohort study, and the time at which the demonstration of a therapeutic response was recorded was 6 months from the start of MTX. Patients who gave informed consent for study participation were generally elderly, and DAS28 was not calculated for all at the start of MTX administration, which also decreased the sample size when they were excluded from the analysis. We could not determine the relative contributions of these polymorphisms to the need for a higher MTX dose because cost-benefit analysis taking into account the time and financial resources needed for genotype determination was not performed. Because MTX is relatively inexpensive, it is likely inferior in cost-effectiveness terms when compared with the precision medicine regimen of MTX administration currently being introduced into clinical practice. In addition, this study did not involve pharmacokinetic and pharmacodynamics (PK–PD) analyses, and thus we were not able to pinpoint the precise difference in MTX dose between responders and nonresponders before prescribing the drug. Therefore, in future precision medicine, it will be necessary to clarify the exact difference in MTX dose between responders and nonresponders by performing cost-effectiveness and PK–PD analyses. As a preliminary study, we examined the factors predicting the therapeutic response to MTX in Japanese patients with RA. MTX dosage adjustment in this study was based on the EULAR criteria and the occurrence of side effects as assessed by the treating rheumatologists. Therefore, it should be noted that the maximum dose of MTX was not prescribed for all patients, which may have affected the results. Because we aimed to clarify factors affecting the therapeutic response to MTX, those contributing to the development of side effects were not investigated. When using genetic information in routine clinical practice, it is necessary to evaluate efficacy and safety simultaneously, and this will be included in future investigations. Only 2 patients used steroids at a maximum dose of 5 mg, but that did not appear to have a major effect on the therapeutic response to MTX. We did not consider organic anion transporters (OAT) 1 and OAT3 gene polymorphisms.

CONCLUSION

The results showed that the therapeutic response to MTX in Japanese RA patients in this study was associated with the SLC19A1 80G>A A/A and GGH 16T>C C/T and CC genotypes. These results could potentially be utilized to formulate therapeutic strategies for patients as form of tailored precision treatment prior to prescribing MTX.

Acknowledgments

We are grateful to Ms. Harumi Kondo and research nurses of the Division of Rheumatology, Department of Internal Medicine, Keio University School of Medicine, Tokyo, Japan, for helpful assistance in collecting patient information.

Conflict of Interest

YK has received lecture fees from Abbvie, Eisai Pharmaceutical, Chugai Pharmaceutical, Bristol Myers Squibb, Astellas Pharmaceutical, Mitsubishi Tanabe Pharma Corporation, Pfizer, Janssen, and UCB. KY has received lecture fees from Astellas Pharma, Asahi Kasei, Takeda Pharmaceutical, Bristol Myers Squibb, MSD LLC, Chugai Pharmaceutical, Mitsubishi Tanabe Pharma Corporation, and DS Pharma Biomedical Co., Ltd. KY has received consulting fees and speaking fees from Pfizer, Chugai Pharma, Mitsubishi-Tanabe Pharma, Takeda Industrial Pharma, GlaxoSmithkline, Nippon Shinyaku, Eli Lilly, Janssen Pharma, Eisai Pharma, Astellas Pharma, and Acterlion Pharmaceuticals. TT has received research grants and lecture fees from Abbvie, Astra Zeneca, Bristol Myers Squibb, Chugai Pharmaceutical, Eisai Pharmaceutical, Janssen Pharmaceutical, Mitsubishi Tanabe Pharma Corporation, Novartis, Takeda Pharmaceutical, Abbott Japan Co., Ltd., Astellas Pharma, Ltd., Daiichi Sankyo, Pfizer, Sanofi–Aventis, Santen Pharmaceutical, Teijin Pharma Ltd., Asahikasei Pharma Corp., SymBio Pharmaceuticals Ltd., Celtrion, Nipponkayaku Co., Ltd., Eli Lilly Japan K.K., and Taisho Toyama Pharmaceutical. JH, MS, JM, MM, and MH declare no conflict of interest.

REFERENCES

• 1) Scott DL, Wolfe F, Huizinga TW. Rheumatoid arthritis. Lancet, 376, 1094–1108 (2010).
• 2) McInnes IB, Schett G. The pathogenesis of rheumatoid arthritis. N. Engl. J. Med., 365, 2205–2219 (2011).
• 3) Jones G, Halbert J, Crotty M, Shanahan EM, Batterham M, Ahern M. The effect of treatment on radiological progression in rheumatoid arthritis: a systematic review of randomized placebo-controlled trials. Rheumatology, 42, 6–13 (2003).
• 4) Pincus T, Ferraccioli G, Sokka T, Larsen A, Rau R, Kushner I, Wolfe F. Evidence from clinical trials and long-term observational studies that disease-modifying anti-rheumatic drugs slow radiographic progression in rheumatoid arthritis: updating a 1983 review. Rheumatology, 41, 1346–1356 (2002).
• 5) Weinblatt ME, Kaplan H, Germain BF, Block S, Solomon SD, Merriman RC, Wolfe F, Wall B, Anderson L, Gall E, Torretti D, Weissman B. Methotrexate in rheumatoid arthritis. A five-year prospective multicenter study. Arthritis Rheum., 37, 1492–1498 (1994).
• 6) Walker AM, Funch D, Dreyer NA, Tolman KG, Kremer JM, Alarcón GS, Lee RG, Weinblatt ME. Determinants of serious liver disease among patients receiving low-dose methotrexate for rheumatoid arthritis. Arthritis Rheum., 36, 329–335 (1993).
• 7) Aletaha D, Smolen JS. The rheumatoid arthritis patient in the clinic: comparing more than 1,300 consecutive DMARD courses. Rheumatology (Oxford), 41, 1367–1374 (2002).
• 8) Maetzel A, Wong A, Strand V, Tugwell P, Wells G, Bombardier C. Meta-analysis of treatment termination rates among rheumatoid arthritis patients receiving disease-modifying anti-rheumatic drugs. Rheumatology, 39, 975–981 (2000).
• 9) Matherly LH, Goldman DI. Membrane transport of folates. Vitam. Horm., 66, 403–456 (2003).
• 10) Chabner BA, Allegra CJ, Curt GA, Clendeninn NJ, Baram J, Koizumi S, Drake JC, Jolivet J. Polyglutamation of methotrexate. Is methotrexate a prodrug? J. Clin. Invest., 76, 907–912 (1985).
• 11) Kremer JM. Toward a better understanding of methotrexate. Arthritis Rheum., 50, 1370–1382 (2004).
• 12) Gorlick R, Goker E, Trippett T, Waltham M, Banerjee D, Bertino JR. Intrinsic and acquired resistance to methotrexate in acute leukemia. N. Engl. J. Med., 335, 1041–1048 (1996).
• 13) Galivan J. Evidence for the cytotoxic activity of polyglutamate derivatives of methotrexate. Mol. Pharmacol., 17, 105–110 (1980).
• 14) Cheng Q, Wu B, Kager L, Panetta JC, Zheng J, Pui CH, Relling MV, Evans WE. A substrate specific functional polymorphism of human γ-glutamyl hydrolase alters catalytic activity and methotrexate polyglutamate accumulation in acute lymphoblastic leukemia cells. Pharmacogenetics, 14, 557–567 (2004).
• 15) Galivan J, Ryan TJ, Chave K, Rhee M, Yao R, Yin D. Glutamyl hydrolase: pharmacological role and enzymatic characterization. Pharmacol. Ther., 85, 207–215 (2000).
• 16) Rhee MS, Lindau-Shepard B, Chave KJ, Galivan J, Ryan TJ. Characterization of human cellular gamma-glutamyl hydrolase. Mol. Pharmacol., 53, 1040–1046 (1998).
• 17) Yao R, Schneider E, Ryan TJ, Galivan J. Human gammaglutamyl hydrolase: cloning and characterization of the enzyme expressed in vitro. Proc. Natl. Acad. Sci. U.S.A., 93, 10134–10138 (1996).
• 18) Cutolo M, Sulli A, Pizzorni C, Seriolo B, Straub RH. Anti-inflammatory mechanisms of methotrexate in rheumatoid arthritis. Ann. Rheum. Dis., 60, 729–735 (2001).
• 19) Shoda H, Inokuma S, Yajima N, Tanaka Y, Oobayashi T, Setoguchi K. Higher maximal serum concentration of methotrexate predicts the incidence of adverse reactions in Japanese rheumatoid arthritis patients. Mod. Rheumatol., 17, 311–316 (2007).
• 20) Kozloski GD, De Vito JM, Kisicki JC, Johnson JB. The effect of food on the absorption of methotrexate sodium tablets in healthy volunteers. Arthritis Rheum., 35, 761–764 (1992).
• 21) Hamilton RA, Kremer JM. The effects of food on methotrexate absorption. J. Rheumatol., 22, 630–632 (1995).
• 22) Kameda H, Amano K, Sekiguchi N, Takei H, Ogawa H, Nagasawa H, Takeuchi T. Factors predicting the response to low-dose methotrexate therapy in patients with rheumatoid arthritis: a better response in male patients. Mod. Rheumatol., 14, 442–446 (2004).
• 23) Yamanaka H, Inoue E, Tanaka E, Nakajima A, Taniguchi A, Terai C, Hara M, Tomatsu T, Kamatani N. Influence of methotrexate dose on its efficacy and safety in rheumatoid arthritis patients: evidence based on the variety of prescribing approaches among practicing Japanese rheumatologists in a single institute-based large observational cohort (IORRA). Mod. Rheumatol., 17, 98–105 (2007).
• 24) Kudo-Tanaka E, Shimizu T, Nii T, Teshigawara S, Yoshimura M, Watanabe A, Tsuji S, Tsuboi H, Hirao M, Yura A, Harada Y, Sueishi M, Suenaga Y, Chiba N, Tonai T, Saisho K, Ogata A, Matsushita M, Hashimoto J, Ohshima S, Tohma S, Saeki Y. Early therapeutic intervention with methotrexate prevents the development of rheumatoid arthritis in patients with recent-onset undifferentiated arthritis: A prospective cohort study. Mod. Rheumatol., 25, 831–836 (2015).
• 25) Hiraga Y, Yuhki Y, Itoh K, Tadano K, Takahashi Y, Mukai M. Pharmacokinetics and efficacy of low-dose methotrexate in patients with rheumatoid arthritis. Mod. Rheumatol., 14, 135–142 (2004).
• 26) Murosaki T, Nagatani K, Sato T, Akiyama Y, Ushijima K, Lefor AK, Fujimura A, Minota S. Prediction of the therapeutic response to methotrexate at 24 weeks by methotrexate-polyglutamates concentration in erythrocytes at 8 weeks in patients with rheumatoid arthritis. Mod. Rheumatol., 27, 411–416 (2017).
• 27) Tazoe Y, Hayashi H, Tsuboi S, Shioura T, Matsuyama T, Yamada H, Hirai K, Tsuji D, Inoue K, Sugiyama T, Itoh K. Reduced folate carrier 1 gene expression levels are correlated with methotrexate efficacy in Japanese patients with rheumatoid arthritis. Drug Metab. Pharmacokinet., 30, 227–230 (2015).
• 28) Hayashi H, Fujimaki C, Daimon T, Tsuboi S, Matsuyama T, Itoh K. Genetic polymorphisms in folate pathway enzymes as a possible marker for predicting the outcome of methotrexate therapy in Japanese patients with rheumatoid arthritis. J. Clin. Pharm. Ther., 34, 355–361 (2009).
• 29) Kato T, Hamada A, Mori S, Saito H. Genetic polymorphisms in metabolic and cellular transport pathway of methotrexate impact clinical outcome of methotrexate monotherapy in Japanese patients with rheumatoid arthritis. Drug Metab. Pharmacokinet., 27, 192–199 (2012).
• 30) Moya P, Salazar J, Arranz MJ, Díaz-Torné C, del Río E, Casademont J, Corominas H, Baiget M. Methotrexate pharmacokinetic genetic variants are associated with outcome in rheumatoid arthritis patients. Pharmacogenomics, 17, 25–29 (2016).
• 31) Zhu H, Deng FY, Mo XB, Qiu YH, Lei SF. Pharmacogenetics and pharmacogenomics for rheumatoid arthritis responsiveness to methotrexate treatment: the 2013 update. Pharmacogenomics, 15, 551–566 (2014).
• 32) Arnett FC, Edworthy SM, Bloch DA, Mcshane D, Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS, Medsger TA Jr, Mitchell DM, Neustadt DH, Pinals RS, Schaller JG, Sharp JT, Wilder RL, Hunder GG. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum., 31, 315–324 (1988).
• 33) Aletaha D, Neogi T, Silman AJ, Funovits J, Felson DT, Bingham CO 3rd, Birnbaum NS, Burmester GR, Bykerk VP, Cohen MD, Combe B, Costenbader KH, Dougados M, Emery P, Ferraccioli G, Hazes JM, Hobbs K, Huizinga TW, Kavanaugh A, Kay J, Kvien TK, Laing T, Mease P, Ménard HA, Moreland LW, Naden RL, Pincus T, Smolen JS, Stanislawska-Biernat E, Symmons D, Tak PP, Upchurch KS, Vencovský J, Wolfe F, Hawker G. 2010 Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Arthritis Rheum., 62, 2569–2581 (2010).
• 34) Hishida A, Matsuo K, Hamajima N, Ito H, Ogura M, Kagami Y, Taji H, Morishima Y, Emi N, Tajima K. Associations between polymorphisms in the thymidylate synthase and serine hydroxymethyltransferase genes and susceptibility to malignant lymphoma. Haematologica, 88, 159–166 (2003).
• 35) Fujii T, Ota M, Hori H, Sasayama D, Hattori K, Teraishi T, Yamamoto N, Hashikura M, Tatsumi M, Higuchi T, Kunugi H. Association between the functional polymorphism (C3435T) of the gene encoding P-glycoprotein (ABCB1) and major depressive disorder in the Japanese population. J. Psychiatr. Res., 46, 555–559 (2012).
• 36) Fukunaga K, Nakagawa H, Ishikawa T, Kubo M, Mushiroda T. ABCB1 polymorphism is associated with atorvastatin-induced liver injury in Japanese population. BMC Genet., 17, 79 (2016).
• 37) Drozdzik M, Rudas T, Pawlik A, Gornik W, Kurzawski M, Herczynska M. Reduced folate carrier-1 80G>A polymorphism affects methotrexate treatment outcome in rheumatoid arthritis. Pharmacogenomics J., 7, 404–407 (2007).
• 38) Kung TN, Dennis J, Ma Y, Xie G, Bykerk V, Pope J, Thorne C, Keystone E, Siminovitch KA, Gagnon F. SLC19A11 80G>A is a genetic determinant of methotrexate efficacy in rheumatoid arthritis: a human genome epidemiologic review and meta-analysis of observational studies. Arthritis Rheumatol., 66, 1111–1120 (2014).
• 39) Baslund B, Gregers J, Nielsen CH. Reduced folate carrier polymorphism determines methotrexate uptake by B cells and CD4+ T cells. Rheumatology, 47, 451–453 (2008).
• 40) van der Straaten RJ, Wessels JA, de Vries-Bouwstra JK, Goekoop-Ruiterman YP, Allaart CF, Bogaartz J, Tiller M, Huizinga TW, Guchelaar HJ. Exploratory analysis of four polymorphisms in human GGH and FPGS genes and their effect in methotrexate-treated rheumatoid arthritis patients. Pharmacogenomics, 8, 141–150 (2007).
• 41) Takahashi C, Kaneko Y, Okano Y, Taguchi H, Oshima H, Izumi K, Yamaoka K, Takeuchi T. Association of erythrocyte methotrexate-polyglutamate levels with the efficacy and hepatotoxicity of methotrexate in patients with rheumatoid arthritis: a 76-week prospective study. RMD Open, 3, e000363 (2017).
• 42) Stranzl T, Wolf J, Leeb BF, Smolen JS, Pirker R, Filipits M. Expression of folylpolyglutamyl synthetase predicts poor response to methotrexate therapy in patients with rheumatoid arthritis. Clin. Exp. Rheumatol., 21, 27–32 (2003).
• 43) Yamamoto T, Shikano K, Nanki T, Kawai S. Folylpolyglutamate synthase is a major determinant of intracellular methotrexate polyglutamates in patients with rheumatoid arthritis. Sci. Rep., 6, 35615 (2016).
• 44) James HM, Gillis D, Hissaria P, Lester S, Somogyi AA, Cleland LG, Proudman SM. Common polymorphisms in the folate pathway predict efficacy of combination regimens containing methotrexate and sulfasalazine in early rheumatoid arthritis. J. Rheumatol., 35, 562–571 (2008).
• 45) Lima A, Seabra V, Bernardes M, Azevedo R, Sousa H, Medeiros R. Role of key TYMS polymorphisms on methotrexate therapeutic outcome in Portuguese rheumatoid arthritis patients. PLoS ONE, 9, e108165 (2014).
• 46) Mandola MV, Stoehlmacher J, Muller-Weeks S, Cesarone G, Yu MC, Lenz HJ, Ladner RD. A novel single nucleotide polymorphism within the 5′ tandem repeat polymorphism of the thymidylate synthase gene abolishes USF-1 binding and alters transcriptional activity. Cancer Res., 63, 2898–2904 (2003).
• 47) Corre S, Galibert MD. Upstream stimulating factors: highly versatile stress-responsive transcription factors. Pigment Cell Res., 18, 337–348 (2005).
• 48) Xiao H, Xu J, Zhou X, Stankovich J, Pan F, Zhang Z, Xu S, Lian L, Ding C. Associations between the genetic polymorphisms of MTHFR and outcomes of methotrexate treatment in rheumatoid arthritis. Clin. Exp. Rheumatol., 28, 728–733 (2010).