A polymorphism and methotrexate efficacy or methotrexate related-toxicity in rheumatoid arthritis: A meta-analysis

International Immunopharmacology 38 (2016) 8–15

Contents lists available at ScienceDirect

International Immunopharmacology journal homepage: www.elsevier.com/locate/intimp

The association between reduced folate carrier-1 gene 80G/A polymorphism and methotrexate efficacy or methotrexate related-toxicity in rheumatoid arthritis: A meta-analysis XiaoBing Li a, MingCai Hu b, WanPing Li a, Li Gu a, MeiJuan Chen a, HuiHua Ding c, Kamala Vanarsa c, Yong Du c,⁎ a b c

College of Pharmacy, Southwest Medical University, Sichuan 646000, China Division of Pharmacy, The Affiliated Hospital of Southwest Medical University, Sichuan 646000, China Department of Biomedical Engineering, University of Houston, Houston, TX, USA

a r t i c l e

i n f o

Article history: Received 22 April 2016 Received in revised form 16 May 2016 Accepted 16 May 2016 Available online xxxx Keywords: Methotrexate Reduced folate carrier 1 Rheumatoid arthritis 80G/A polymorphism Efficacy Toxicity Meta-analysis

a b s t r a c t Methotrexate (MTX), the most commonly used anti-rheumatic drug against RA, enters the cell via the action of the reduced folate carrier 1(RFC1). A major polymorphism of the RFC1 gene, 80G/A, has been reported to influence the activity of RFC1, resulting in variable intracellular MTX-polyglutamate (MTX-PG) levels. However, the association studies addressing the RFC1 80G/A polymorphism and MTX efficacy or toxicity in Rheumatoid arthritis (RA) has yielded conflicting results. In the present meta-analysis, we aimed to evaluate the association between the RFC1 80G/A polymorphism and MTX efficacy or toxicity in RA patients. A total 17 studies met our inclusion criteria. Among them, 12 studies with 2049 subjects reported the association between the RFC1 80G/ A and MTX response, and 12 studies involving 2627 subjects were on MTX-related toxicity. Meta-analysis revealed significant association between RFC1 80G/A polymorphism and MTX efficacy (odds ratio (OR) for the A allele = 1.29, 95% confidence interval (CI) 1.05–1.67, P = 0.02; for AA genotype: OR = 1.49, 95%CI 1.17– 1.907, P = 0.001). However, no association could be detected in the analysis of MTX-related toxicity. Stratification by ethnic population also indicated an association between this polymorphism and MTX efficacy in Asian group (P = 0.002 for A allele; P = 0.003 for AA genotype), but not in the Caucasian group (P = 0.15 for A allele; P = 0.05 for AA genotype). In both Asian and Caucasian sub-groups, no influence of the RFC1 80G/A polymorphism on MTX toxicity can be detected. In conclusion, the RFC1 G80A polymorphism is associated with responsiveness to MTX therapy, but may not be associated with MTX toxicity in RA patients. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Rheumatoid arthritis (RA) is a chronic, systemic autoimmune condition which manifests itself in multiple joints of the body. Patients with RA commonly present with joint pain and stiffness, and progressive functional disability. As the most common inflammatory arthritis, RA has a substantial negative impact on all aspects of life quality of patients. It accounts for 0.84% of all disability-adjusted life years lost in Europe, and 0.64% in the USA [1,2]. Hence, it poses significant economic and social burden on patients and society. Furthermore, according to recent epidemiological studies, the RA-associated socioeconomic burden is expected to grow, as the population with arthritis-attributable activity limitations is projected to increase to 25 million in the USA by 2030 [3]. Methotrexate (MTX), the conventional disease-modifying antirheumatic drug (DMARD), is the first-line anchor drug for treating RA. ⁎ Corresponding author at: Department of Biomedical Engineering, University of Houston, 3605 Cullen Blvd, Houston, TX 77204, United States. E-mail address: [email protected] (Y. Du).

http://dx.doi.org/10.1016/j.intimp.2016.05.012 1567-5769/© 2016 Elsevier B.V. All rights reserved.

However, MTX usage displays significant interpatient variability and a broad spectrum of unpredictable adverse drug reactions (ADRs). Approximately, only 35–50% of patients experience good clinical response, and 30% develop ADRs and have to discontinue MTX treatment [4,5]. Although various factors may contribute to the variability associated with MTX response and/or ADRs, it is clear that the polymorphisms in genes coding for MTX transporters and metabolizing enzymes play a critical role [6]. MTX enters cells through the action of the reduced folate carrier (RFC1), also known as the solute carrier family 19 folate transporter member 1 (SLC19A1). RFC1 is a dominant, bidirectional, cell membrane transporter, transferring hydrophilic folates across the cell membrane using negatively charged glutamate residue of folate. Therefore, the RFC1 gene/protein expression level will influence how cells uptake MTX, then impact the efficacy and toxicity of MTX treatment in many ways. In supporting this, Tazoe Y et al. recently reported that RFC1 gene expression levels are correlated with MTX efficacy in Japanese patients with RA, as disease activity scores were lower for patients with higher RFC1 mRNA levels [7]. Moreover, it is believed that some functional mutations in RFC1 gene

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

may alter the transporter structure and such impact MTX uptake, modify MTX bioavailability and, consequently, influence the MTX efficacy [8–11]. A common polymorphism at position 80 in exon 2 of RFC1 (rs1051266), which results in a change from guanine (G) to adenine (A), confers the capacity to determine the efficiency of MTX uptake by B cells and CD4 + T cells [12]. The association between this polymorphism and the clinical outcome in patients with acute lymphoblastic leukemia (ALL) has also been reported in series of studies [9,11,13–15]. Dervieux et al. first investigated the contribution of this polymorphism to the MTX efficacy in patients with RA. They reported that RA patients with AA genotype had an approximate 2-fold higher frequency of MTX-polyglutamate (MTX-PG) levels. More importantly, RA patients with AA genotype displayed lower disease severity when compared with those with the RFC1 GG and GA genotypes [16,17]. Following this, the association between RFC1 80G/A and MTX efficacy or ADRs in patients with RA has been extensively studied [18–35,40–46]. However, these studies have shown contradictory results. For instance, Hayashi H et al. reported that patients carrying the A allele at 80G/A have higher intracellular MTX uptake and better response to MTX treatment [25], which was supported by several independent studies [19,21,28], but not others [18,20,22–24,26,27]. Regarding the MTX-related toxicity, Lima et al. found an association between this polymorphism and MTX-related gastrointestinal (GI) toxicity in Portuguese RA patients [35], but not the overall toxicity [32]. The discrepancies of these findings might be owing to the difference in the study design (cross-sectional vs. longitudinal), the study populations (Asian vs. Caucasian), the duration of MTX treatment, various response or ADR judgment criteria, as well as co-treatment with other DMARDs [8]. In the present study, we conducted a meta-analysis to examine the association between RFC1 80G/A polymorphism and MTX efficacy or MTX-related ADRs in RA patients receiving MTX treatment. Our results support the notion that this polymorphism is associated with MTX efficacy, but not MTX related-ADRs, indicating RFC1 80G/A can be used as a biomarker to predict the MTX response.

9

2. Materials and methods 2.1. Identification of eligible studies and data extraction We sought studies in which associations between the RFC1 80G/A gene polymorphism and the efficacy or toxicity of MTX had been examined in adult populations with RA. A combination of keywords, Medical Subject Heading (MeSH) terms, and text words, including “Rheumatoid Arthritis,” “Methotrexate,” “RFC1 80G/A”, “rs1051266,” “SLC19A1,” “pharmacogenetics,” and “genetic association,” has been used to research and identify related literature through PubMed, EMBASE and Cochrane Library Citations from Jan./1990 to Jan./2016. Abstracts from the American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR) annual meetings (2009 to 2015) were also reviewed to identify additional studies that were not indexed in the electronic databases. A manual search for additional relevant studies using the references from these retrieved articles was also performed. The language of origin was limited to English. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for a systematic review and meta-analysis. Each retrieved study was read in its entirety to assess the appropriateness for inclusion in the meta-analysis. The following information was extracted by two reviewers (Li and Hu): author, year of publication, studied population, disease duration, MTX dose, response criteria, the length of follow-up period, Hardy–Weinberg equilibrium, and genotype distribution of RFC1 80G/A polymorphism. 2.2. Inclusion and exclusion criteria The following inclusion criteria have been used: (1) Evaluation of the associations between the RFC1 80G/A gene polymorphism and the efficacy or toxicity of MTX treatment in adult patients with RA; (2) case-control study design; (3) detailed genotype data could be acquired to calculate the odds ratios (ORs) and 95% confidence intervals (CIs); (4) per-reviewed publications in English. Exclusion criteria include (1) duplication of previous publications; (2) comment, review, editorial

Fig. 1. Flow chart of systemic search and study selection.

10

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

Table 1 Characteristics of studies included in the MTX efficacy meta-analysis. Patient Author year (ref) number Population/ethnicity Study design Takatori R et al. 105 (2006) [18] Drozdzik et al. 174 (2007) [19] Chatzikyriakidou 54 et al (2007) [20]

Poland/Caucasian

Retrospective Responders: patients with the last MTX maintenance dosage b6 mg/week Prospective ACR 20% response criteria in 12 months

Greece/Caucasian

Retrospective DAS28 , ACR 20% and ACR 50%

98

England/Caucasian

Prospective

213

Slovenia/ Caucasian

255

USA/Caucasian

187 170

New Zealand/Caucasian Japanese

Retrospective Responders: patients with a mean DAS28 above 2.6 and below 3.2 and with a reduction in DAS28 N 1.2 Retrospective EULAR response criteria or physician's assessment of patient's response (VAS) Prospective Responders: a reduction in DAS N 1.2 and low disease activity defined as DAS28 b 3.2 Retrospective DAS 28

Owen et al. (2013) [26]

231

UK/Caucasian

Kung, et al. (2014) [27] Ghodke-Puranik et al. (2015) [28] Muralidharan et al. (2016) [29]

120

Canadian/Caucasian

Retrospective Responder: physician statement of good response plus a stable dose of MTX for at least 6 months, with an ESR b 20 and/or normal CRP Retrospective ND

217

Indian

225

Indian

James et al. (2008) [21] Bohanec et al. (2008) [22] Derivieux et al. (2009) [23] Stamp et al. (2010) [24] Hayashi et al. (2013) [25]

Japanese

Judgment criteria

MTX dose (mg/week)

Main findings

ND

NS

7.5–15 11.57 ± 2.8 for MTX only 13.78 ± 3.08 for MTX+ DMARDs 15 (10-25)

Responders

Non-responders

GG GA AA GG

GA

AA

14

41

17

6

18

9

A↑ 8 response NS 18

28

22

29

63

24

21

11

3

1

0

A↑ 22 response 10.0(10.0–12.5) NS 49

41

20

9

5

1

82

49

11

16

6

15 (10-15)

NS

35

79

22

43

58

18

15 (5-25)

NS

41

48

22

24

40

12

6.0(2.0–10.0) for MTX, 8.0(4.0–12.0) for MTX +

A↑ 14 response

45

30

23

45

13

DMARDs ND

NS

50

63

26

23

50

19

ND

NS

33

49

19

5

10

4

Retrospective ACR50

17.5 (11.25–17.5)

A↑ 13 response

25

11

74

70

24

Prospective

16.75 ± 4

NS

64

21

35

59

12

EULAR (DAS28 at 12 months)

EULAR response criteria at 16 weeks

34

Note: 1) ND: no data or data cannot be collected; 2) NS: no association; 3) DMARDs: disease-modifying anti-rheumatic drugs; 4) for MTX dose, the data are presented as mean ± standard deviation, or mean (minimum–maximum); 5) DAS28: 28-joint count Disease Activity Score; 6): ACR 20%: American College of Rheumatology (ACR) 20% improvement response criteria; 7) EULAR: European League Against Rheumatism; 8) CRP: C-reactive protein; 9) ESR: erythrocyte sedimentation rate.

and conference abstract; (3) study with no detailed genotype data; (4) Non-English publications. Each study was screened in duplicate by two independent reviewers (Li and Du) per the guidelines of the Human Genomic Epidemiology (HuGE) Review Handbook. Of note, for studies of MTX efficacy, all measures of disease activity were accepted, which mainly included the 44-joint count Disease Activity Score (DAS44) or the 28-joint count DAS (DAS28) or Physician's global assessment of disease (VAS score) and the ACR 20% or ACR 50% improvement response criteria (ACR20 or ACR50) [27]. For studies of MTX toxicity, any adverse event was accepted. These adverse events can be identified by clinical history, physician exam as well as lab results. 2.3. Evaluation of statistical associations Hardy–Weinberg equilibrium (HWE) was accessed for each study by Chi-square test in response groups, and P b 0.05 was considered a significant departure from HWE. Statistical heterogeneity was assessed using Cochran's Q test. This statistic was complemented with the I2 statistic, which quantifies the proportion of the total variation across studies that is due to heterogeneity rather than chance. The I2 values of 25%, 50%, and 75% corresponded to low, medium, and high between-study heterogeneity and Cochran's Q statistic P b 0.05 suggested no statistically significant heterogeneity [27]. Then, based on the heterogeneity, either fixed-effect or random-effect model was selected to pool the effect sizes. In additional, publication bias was assessed using the Egger's test and expressed graphically using funnel plots, and the P value of Egger's test b0.05 with an asymmetric plot was considered a significant publication bias [36]. We also conducted subgroup analysis

by ethnic group, as allele or genotype frequencies of RFC1 80G/A polymorphism vary with race [37–39]. Review Manager Software (Version 5.3, Copenhagen, Denmark) from the Cochrane Collaboration was used for data analysis. 3. Results 3.1. Studies included in the meta-analysis A total of 53 per-review publications were acquired from PubMed, EMBASE and Cochrane Library citation databases (PubMed: 23; EMBASE: 30; Cochrane Library citation: 0). After initial title and abstract review, 26 articles have been excluded, and 27 were selected for fulltext review [8,16–35,40–45]. Among these 27 studies, 6 studies are not case-control design and multiple variants have been analyzed together, hence, no detailed genotype data could be retrieved [8,16,17, 40–42]. Three case-control designed studies didn't report the complete genotype data (43–45). One study was excluded to avoid inclusion of duplicated data as the same patients and controls seemed to be reported by same authors in both studies [32,35]. Therefore, 17 relevant studies addressing the relationship between RFC1 80G/A polymorphism and the efficacy or toxicity of MTX in RA patients were included in our meta-analysis. The literature review process is shown in Fig. 1. Among these 17 studies, 12 examined the MTX efficacy, with a total population of 1187 responders and 862 non-responders.12 studies investigated the MTX toxicity on RA patients, with 978 ADRs and 1649 cases without ADRs. Tables 1 and 2 summarized the characteristics of these studies recruited in the analysis, such as sample size, population, study results, and genotype distribution. The overall findings of these

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

11

Table 2 Characteristics of studies included in the MTX toxicity meta-analysis.

Author year (ref) Takatori et al. (2006) [18] Drozdzik et al. (2007) [19] Bohanec et al. (2008) [22] Derivieux et al. (2009) [23] Stamp et al. (2010) [24] Cáliz R et al. (2012) [30] Plaza-plaza et al. (2012) [31]

Patient Number

Population/ethnicity

Study design

124

Japanese

174

Poland/Caucasian

Retrospective Physician examination and lab results Prospective Lab results, ALT/AST N10UI

150

Slovenia/ Caucasian

158 187

Non-ADE group

ADE group

MTX dose (mg/week)

Main findings

ND

NS

10

29

9

17

39 20

7.5–15

NS

1

5

4

42

89 38

Retrospective Clinical history and lab results

ND

35

51

27

5

18 14

USA/Caucasian

Retrospective Clinical history and lab results

ND

GG ↑ toxicity NS

11

20

6

37

68 16

Prospective

ND

NS

45

69

22

20

19 12

468

New Zealand/Caucasian Spanish/Caucasian

Retrospective Lab results: ALT/AST N10UI

10–25

NS

23

39

22

52

Spanish/Caucasian

Retrospective Clinical history and lab results

10.0 ± 2.10

NS

11

10

5

7

Judgment criteria

A standardized questionnaire ADRs

GG GA AA GG

GA

AA

109 187 88 16

4

15.0 GG ↑ (2.5–25.0) toxicity 15.0(12.5–25) NS

28

33

16

52

58 46

43

59

25

29

62 19

332

Portuguese/Caucasian Retrospective Clinical history and lab results classified in SOC Poland/Caucasian Retrospective Patient report, physician examination and lab results Indian Retrospective Clinical history and lab results

15.0 ± 3.9

NS

28

74

68

22

75 55

194 327

Spanish/Caucasian Indian

12.33 ± 4.10 ND

NS NS

23 27

50 32

10 8

33 60 18 75 138 47

Lima et al. (2014) [32]

233

Świerkot et al. (2015) [33] Ghodke-Puranik et al. (2015) [28] Moya et al. (2016) [34] Muralidharan et al. (2016) [29]

240

Retrospective Clinical history and lab results Prospective Lab results: ALT/AST N10UI

Note: 1) ADR: adverse drug reactions; 2) ND: no data or data cannot be collected; 3) NS: No association; 4) ALT: alanine aminotransferase; 5) AST: aspartate aminotransferase; 6) SOC: System Organ Class. 7) for MTX dose, the data are presented as mean ± standard deviation, or mean (minimum–maximum).

studies were very controversial. Only 4 studies reported a positive association between RFC1 80G/A polymorphism and MTX response, whereas, only two studies indicated RA patients with GG genotype had a higher risk of MTX-associated toxicity. 3.2. The association of RFC1 80G/A polymorphism and MTX efficacy in RA patient A total of 2049 RA cases were reported in these 12 studies recruited in our meta-analysis. Among them, 1187 patients showed a good response, with an average response rate of 57.9% (range from 22.5% to 92.6% based on each report). The frequency of A allele in responder patients was higher than that of the non-responder group (47.4% vs. 41.7%; χ2 = 144.2, P b 0.0001). Similarly, the distribution of AA

genotype in responder group was significantly higher than that of the non-responder group (22.7% vs. 16.5%; χ2 = 11.85, P = 0.0006). These results supported that RA patients carrying A allele or AA genotype had a better response to MTX treatment. Our meta-analysis results were shown in Table 3 and Fig. 2. Similar to a previous report [27], we also found a significant association between RFC1 80G/A variant and MTX efficacy in RA patients. A fixed-effect model was conducted for the recessive model analysis (AA vs. AG+ GG), as no between-study heterogeneity was detected (I2 = 1%, P = 0.43). The odd ratio (OR) estimated for the MTX efficacy increased by 49% for RA patients carrying AA genotype, compared to those with AG + GG genotype (95% CI 1.17–1.90). For the allelic model, the overall OR was 1.29 for patients with A allele (95% CI 1.05–1.67), compared to patients carrying G allele, with a random-effect model analysis (P =

Table 3 Association between RFC1 80G/A polymorphism and MTX efficacy in RA patients.

Sample size

Test of heterogeneity

Test of association

Outcome

No. of Included studies

NO. of responder

No. of non-responder

P

Model

Z

P

OR (95%CI)

Overall studies A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

12 12 12 12

2374 1187 1187 1187

1724 862 862 862

0.02 0.43 0.03 0.54

Random Fixed Random Fixed

2.39 3.19 1.89 0.34

0.02 0.001 0.06 0.73

1.29(1.05–1.67) 1.49(1.17–1.90) 0.73(0.53–1.01) 0.97(0.80–1.17)

Caucasian A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

8 8 8 8

1716 858 858 858

948 474 474 474

0.02 0.47 0.02 0.35

Random Fixed Random Fixed

1.45 1.99 1.09 0.55

0.15 0.05 0.28 0.58

1.24(0.93–1.67) 1.37(1.00–1.86) 0.78(0.49–1.23) 0.93(0.73–1.19)

Asian A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

4 4 4 4

658 329 329 329

776 388 388 388

0.19 0.30 0.37 0.60

Fixed Fixed Fixed Fixed

3.11 2.65 2.52 0.15

0.002 0.003 0.01 0.72

1.42(1.14–1.77) 1.71(1.15–2.55) 0.63(0.45–0.90) 1.02(0.75–1.40)

Note: 1) OR: odds ratio; 2) 95% CI: 95% confidence interval.

12

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

Fig. 2. Association between RFC1 80G/A and MTX efficacy among all studies identified in a Fixed-effects meta-analysis under the recessive genetic model (AA vs. GG + AG). A) Forest plot. Square data markers represent OD; horizontal lines, the 95% CI with marker size reflecting the statistical weight of the study. A diamond data marker represents the overall OD and 95% CI for the outcome of interest. B) Funnel pot. The shape of the funnel plot indicated no obvious asymmetry.

0.02). These findings indicated that patients with A allele have a better response to MTX treatment than those carrying G allele. No significant association has been found in both dominant, and over-dominant model (P = 0.06 and P = 0.73 respectively, Table 3). To evaluate the race-specific effect, we divided the population by ethnicity. There was no considerable heterogeneity across the Asian studies. Consequently, the fixed-effect model was selected. The overall OR of MTX efficacy for the AA genotype in Asians was 1.71 (95% CI 1.15–2.55, P = 0.003; Table 3). It was 1.42 for the A allele in Asians (95% CI 1.14–1.77, P = 0.002; Table 3). These data indicated that Asian RA patients carrying A allele had better response to MTX treatment. We also identified similar magnitude in the Caucasian sub-group, as the overall OR of MTX efficacy for the AA genotype was 1.37 (95% CI 1.00–1.86), though without statistical significance (P = 0.05, Table 3).

(ranging from 5.75% to 72.7%, based on each report). The distribution of A allele and AA genotype in ADR group was similar as that of the non-ADR group (AA genotype: 22.7% vs. 22.4%, P = 0.93; A allele: 48.6% vs. 52.5%, P = 0.52, respectively), indicating there was no association between this polymorphism and MTX toxicity. As Table 4 and Fig. 3 indicated, no significant heterogeneity has been detected between-studies in the analysis, and fixed-effect model has been used to pool the combined effect. In agreement with the pervious study [27], our meta-analysis also did not find a notable association between RFC1 80G/A polymorphism and MTX Toxicity, no matter which genetic model was applied. Because more evidence and data has been reported since the last report, we could conduct the sub-population analysis. However, in both Caucasian and Asian sub-population, no significant association could be detected (Table 4).

3.3. The relationship between RFC1 80G/A polymorphism and MTX Toxicity in RA patient

4. Discussion

There were 2627 cases in the 12 studies included in our meta-analysis, with 978 patients with ADRs and 1649 cases without ADRs, approximately 37.2% RA patients developing ADRs after MTX treatments

In the present study, we performed a meta-analysis using available data to examine whether RFC1 80G/A polymorphism is related to MTX efficacy or MTX associated-ADRs in RA patients. Our results demonstrated an association between RFC1 80G/A and MTX efficacy, but

Table 4 Association between RFC1 80G/A polymorphism and MTX related-ADRs in RA patients.

Sample size

Test of heterogeneity

Test of association

Outcome

No. of Included studies

NO. of ADRs

No. of non-ADRs

P

Model

Z

P

OR (95%CI)

Overall studies A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

12 12 12 12

1956 978 978 978

3298 1649 1649 1649

0.92 0.37 0.47 0.41

Fixed Fixed Fixed Fixed

1.40 0.91 1.47 0.52

0.16 0.36 0.14 0.60

0.92(0.81–1.04) 0.91(0.74–1.12) 1.16(0.95–1.41) 0.95(0.80–1.14)

Caucasian A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

9 9 9 9

1386 693 693 693

2322 1161 1161 1161

0.03 0.33 0.37 0.34

Random Fixed Fixed Fixed

1.20 0.81 0.87 0.13

0.23 0.42 0.38 0.89

0.87(0.69–1.10) 0.90(0.69–1.17) 1.11(0.88–1.40) 0.98(0.78–1.24)

Asian A vs. G (Per-allele model) AA vs. AG + GG (recessive model) AG + AA vs. GG (dominant model) AG vs. AA + GG (over-dominant model)

3 3 3 3

570 285 285 285

976 488 488 488

0.69 0.26 0.47 0.35

Fixed Fixed Fixed Fixed

0.43 0.44 1.39 0.76

0.67 0.66 0.17 0.45

1.05(0.84–1.31) 0.92(0.65–1.32) 1.30(0.90–1.89) 0.89(0.65–1.21)

Note : 1) ADR: adverse drug reactions; 2) OR: odds ratio; 3) 95% CI: 95% confidence interval.

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

13

Fig. 3. Association between RFC180G/A and MTX related-ADRs among all studies identified in a Fixed-effects meta-analysis under the recessive genetic model (AA vs. GG + AG). A) Forest plot. Square data markers represent OD; horizontal lines, the 95% CI with marker size reflecting the statistical weight of the study. A diamond data marker represents the overall OD and 95% CI for the outcome of interest. B) Funnel pot. The shape of the funnel plot indicated no obvious asymmetry.

not MTX toxicity, indicating that the variant might be a potential predictor for MTX response, but not a useful biomarker for predicting MTX-related toxicity. Over the last decade, the field of pharmacogenomics has attracted much attention, as it may provide a possible explanation for individual differences in drug response and/or side-effects [6,46]. The response to low-dose MTX therapy in RA patients demonstrates wide interpatient variability, with still unknown reasons. It is clear that various MTX transporters and metabolizing enzymes are involved in the MTX metabolism. These genes include RFC, ATP-binding cassette (ABC), Dihydrofolate reductase (DHR), methionine synthase reductase (MTRR), as well as Methylenetetrahydrofolate reductase (MTHFR), etc. Hence, in the last decade, many studies have been conducted to explore whether these genes and their variants are associated with the variability. RFC1 plays a critical role in regulating cells transporting MTX into cells. It is relevant that RFC gene polymorphism and/or the expression of RFC1 protein contribute to the interpatient variability observed in MTX response and MTX- associated toxicity. Indeed, RFC1 80G/A polymorphism is associated with the levels of intracellular MTX-PGs. Once MTX is transported into the cells, it will be converted into several types of MTX-PGs. Increasing evidence demonstrated that Long-chain MTX-PGs are biomarkers of MTX efficacy, as longer chain MTX-PGs are better retained in cells, with longer anti-folate effects than these with shorter chains. Several studies demonstrated that patients with RFC1 80AA genotype had higher MTX-PG3–5 and/ or MTX-PG5 levels in RBC [16,17,25]. These data supported this polymorphism would impact the transportation of MTX into the cells, leading to the different level of MTX-PGs. Likewise, increased RFC1 mRNA expression may increase MTX uptake by immune cells [7,12], and as a result, increase the clinical efficacy and reduce disease activity in RA patients receiving MTX. However, contradictory results on the association between the RFC1 80G/A with the MTX efficacy have been reported in the literature. Our meta-analysis pooled the effect of 12 qualified studies and found RFC1 80G/A variant significantly associated with the MTX efficacy. The odd of MTX efficacy is increased by 49% for these carrying AA genotype, compared to these with AG and GG genotypes. Similarly, the odd of MTX efficacy increased by 29% in patients with A allele, when compared with carriers of G allele. These results are consistent with a previous meta-analysis reported by Kun and colleagues [27], and the recent data observed in ALL patients [13–15]. In the sub-population analysis,

both Caucasian and Asian population demonstrated similar pooled effect size (Table 3). Several concerns raised in integrating the results of the meta-analysis. Although all patients followed the same RA diagnosis criteria, the outcome measurement varied among these studies recruited in our meta-analysis, as multiple evaluation criteria have been used. Only one study from Wessels et al. was randomized control trial (RCT), while others were either retrospective or prospective observation studies [44]. No association could be detected in this RCT study. Moreover, variability in MTX dose and the use of other DMARDs will add inerasable impacts on the outcome, as such, the results of our meta-analysis. In term of MTX toxicity, no significant association between RFC1 80G/A polymorphism and MTX toxicity was observed in our meta-analysis, which was in agreement with a pervious meta-analysis in RA. Our sub-population analysis did not detect any association as well. Similarly, the results of a recent meta- analysis conducted in childhood ALL also revealed that RFC1 80G/A polymorphism was not associated with MTX toxicity [47]. Taken together, this evidence indicated that RFC1 80G/A does not seem to be a good marker of MTX-related toxicity. Similar to the analysis of MTX efficacy, we again need to read the result in caution. The biggest challenge is how to define the MTX-associated toxicity. A broad spectrum of criteria was applied to evaluate the adverse events in each study, which might overestimate the MTX-related toxicity [27]. Heterogeneous in study design was also significant, as some studies only examined the GI toxicity, while most others included all possible events. Therefore, standardization of criteria of toxicity should be considered in the future studies. What's more important, N14 polymorphisms have been identified in RFC1 gene. In Portuguese RA patients, Lima A et al. reported that the overall MTX toxicity was related to RFC1 rs7499 [32], while GI toxicity was associated with RFC1 rs7499, or 80G/A or, rs2838956 respectively [35]. These results implicated that different polymorphisms in the RFC1 gene function diversely and confer different association with various clinical outcomes. Newly identified polymorphism, rs1051296, a functional mutation in the miRNA-binding site of RFC1, has also been found associated with MTX plasma concentration in Chinese ALL patients [10]. Such, more studies are warranted to explore the association between RFC1 gene polymorphisms and MTX-related toxicity. Conflict of interest The authors declare that they have no conflict of interest.

14

X. Li et al. / International Immunopharmacology 38 (2016) 8–15

Funding sources This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors

References [1] T. Uhlig, R.H. Moe, T.K. Kvien, The burden of disease in rheumatoid arthritis, PharmacoEconomics 32 (9) (2014) 841–851. [2] L. Gulácsi, V. Brodszky, P. Baji, H. Kim, S.Y. Kim, Y.Y. Cho, M. Péntek, Biosimilars for the management of rheumatoid arthritis: economic considerations, Expert. Rev. Clin. Immunol. 11 (S1) (2015) S43–S52. [3] C.G. Helmick, D.T. Felson, R.C. Lawrence, S. Gabriel, R. Hirsch, C.K. Kwoh, M.H. Liang, H.M. Kremers, M.D. Mayes, P.A. Merkel, S.R. Pillemer, J.D. Reveille, J.H. Stone, Estimates of the prevalence of arthritis and other rheumatic conditions in the United States: part I, Arthritis Rheum. 58 (1) (2007) 15–25. [4] A. Maetzel, A. Wong, V. Strand, P. Tugwell, G. Wells, C. Bombardier, Meta-analysis of treatment termination rates among rheumatoid arthritis patients receiving diseasemodifying anti-rheumatic drugs, Rheumatology (Oxford) 39 (9) (2000) 975–981. [5] D. Aletaha, J.S. Smolen, The rheumatoid arthritis patient in the clinic: comparing N1.300 consecutive DMARD courses, Rheumatology (Oxford) 41 (12) (2002) 1367–1374. [6] P.I. Burgos, M.I. Danila, J.M. Kelley, L.B. Hughes, S.L. Bridges Jr., Understanding personalized medicine in rheumatoid arthritis: a Clinician's guide to the future, Ther. Adv. Musculoskelet. Dis. 1 (2) (2009) 97–105. [7] Y. Tazoe, H. Hayashi, S. Tsuboi, T. Shioura, T. Matsuyama, H. Yamada, K. Hirai, D. Tsuji, K. Inoue, T. Sugiyama, K. Itoh, Reduced folate carrier 1 gene expression levels are correlated with methotrexate efficacy in Japanese patients with rheumatoid arthritis, Drug Metab. Pharmacokinet. 30 (3) (2015 Jun) 227–230. [8] P. Bohanec Grabar, L.J. Leandro-García, L. Inglada-Pérez, D. Logar, C. RodríguezAntona, V. Dolžan, Genetic variation in the SLC19A1 gene and methotrexate toxicity in rheumatoid arthritis patients, Pharmacogenomics 13 (14) (2012) 1583–1594. [9] C. Laverdière, S. Chiasson, I. Costea, A. Moghrabi, M. Krajinovic, Polymorphism G80A in the reduced folate carrier gene and its relationship to methotrexate plasma levels and outcome of childhood acute lymphoblastic leukemia, Blood 100 (2002) 3832–3834. [10] S.M. Wang, L.L. Sun, W.X. Zeng, W.S. Wu, G.L. Zhang, Effects of a microRNA binding site polymorphism in SLC19A1 on methotrexate concentrations in Chinese children with acute lymphoblastic leukemia, Med. Oncol. 31 (7) (2014) 62. [11] E. Lopez-Lopez, J. Ballesteros, M.A. Piñan, S. de Toledo J, G. de Andoin N, P. GarciaMiguel, A. Navajas, A. Garcia-Orad, Polymorphisms in the methotrexate transport pathway: a new tool for MTX plasma level prediction in pediatric acute lymphoblastic leukemia, Pharmacogenet. Genomics 23 (2) (2013) 53–61. [12] B. Baslund, J. Gregers, C.H. Nielsen, Reduced folate carrier polymorphism determines methotrexate uptake by B cells and CD4+ T cells, Rheumatology (Oxford) 47 (4) (2008) 451–453. [13] P. Chiusolo, S. Giammarco, S. Bellesi, E. Metafuni, N. Piccirillo, D. De Ritis, S. Marietti, S. Federica, L. Laurenti, L. Fianchi, S. Hohaus, L. Giuseppe, S. Sica, The role of MTHFR and RFC1 polymorphisms on toxicity and outcome of adult patients with hematological malignancies treated with high-dose methotrexate followed by leucovorin rescue, Cancer Chemother. Pharmacol. 69 (3) (2012) 691–696. [14] L.J. Ashton, A.J. Gifford, E. Kwan, A. Lingwood, D.T. Lau, G.M. Marshall, M. Haber, M.D. Norris, Reduced folate carrier and methylenetetrahydrofolate reductase gene polymorphisms: associations with clinical outcome in childhood acute lymphoblastic leukemia, Leukemia 23 (7) (2009) 1348–1351. [15] J. Gregers, I.J. Christensen, K. Dalhoff, B. Lausen, H. Schroeder, S. Rosthoej, N. Carlsen, K. Schmiegelow, C. Peterson, The association of reduced folate carrier80GNA polymorphism to outcome in childhood acute lymphoblastic leukemia interacts with chromosome 21 copy number, Blood 115 (23) (2010) 4671–4677. [16] T. Dervieux, D. Furst, D.O. Lein, R. Capps, K. Smith, M. Walsh, J. Kremer, Polyglutamation of methotrexate with common polymorphisms in reduced folate carrier, aminoimidazole carboxamide ribonucleotide transformylase, and thymidylate synthase are associated with methotrexate effects in rheumatoid arthritis, Arthritis Rheum. 50 (9) (2004) 2766–2774. [17] T. Dervieux, D. Furst, D.O. Lein, R. Capps, K. Smith, J. Caldwell, J. Kremer, Pharmacogenetic and metabolite measurements are associated with clinical status in patients with rheumatoid arthritis treated with methotrexate: results of a multicentred cross sectional observational study, Ann. Rheum. Dis. 64 (8) (2005) 1180–1185. [18] R. Takatori, K.A. Takahashi, D. Tokunaga, T. Hojo, M. Fujioka, T. Asano, T. Hirata, Y. Kawahito, Y. Satomi, H. Nishino, T. Tanaka, Y. Hirota, T. Kubo, ABCB1 C3435T polymorphism influences methotrexate sensitivity in rheumatoid arthritis patients, Clin. Exp. Rheumatol. 24 (5) (2006) 546–554. [19] M. Drozdzik, T. Rudas, A. Pawlik, W. Gornik, M. Kurzawski, M. Herczynska, Reduced folate carrier-1 80GNA polymorphism affects methotrexate treatment outcome in rheumatoid arthritis, Pharmacogenomics J. 7 (6) (2007) 404–407. [20] A. Chatzikyriakidou, I. Georgiou, P.V. Voulgari, C.G. Papadopoulos, T. Tzavaras, A.A. Drosos, Transcription regulatory polymorphism −43TNC in the 5′-flanking region of SLC19A1 gene could affect rheumatoid arthritis patient response to methotrexate therapy, Rheumatol. Int. 27 (11) (2007) 1057–1061. [21] H.M. James, D. Gillis, P. Hissaria, S. Lester, A.A. Somogyi, L.G. Cleland, S.M. Proudman, Common polymorphisms in the folate pathway predict efficacy of combination regimens containing methotrexate and sulfasalazine in early rheumatoid arthritis, J. Rheumatol. 35 (4) (2008) 562–571.

[22] P. Bohanec Grabar, D. Logar, B. Lestan, V. Dolzan, Genetic determinants of methotrexate toxicity in rheumatoid arthritis patients: a study of polymorphisms affecting methotrexate transport and folate metabolism, Eur. J. Clin. Pharmacol. 64 (11) (2008) 1057–1068. [23] T. Dervieux, J.A. Wessels, T. van der Straaten, N. Penrod, J.H. Moore, H.J. Guchelaar, J.M. Kremer, Gene-gene interactions in folate and adenosine biosynthesis pathways affect methotrexate efficacy and tolerability in rheumatoid arthritis, Pharmacogenet. Genomics 19 (12) (2009) 935–944. [24] L.K. Stamp, R.L. Roberts, Effect of genetic polymorphisms in the folate pathway on methotrexate therapy in rheumatic diseases, Pharmacogenomics 12 (10) (2011) 1449–1463. [25] H. Hayashi, Y. Tazoe, S. Tsuboi, M. Horino, M. Morishita, T. Arai, M. Ohshima, T. Matsuyama, K. Kosuge, H. Yamada, D. Tsuji, K. Inoue, K. Itoh, A single nucleotide polymorphism of reduced folate carrier 1 predicts methotrexate efficacy in Japanese patients with rheumatoid arthritis, Drug Metab. Pharmacokinet. 28 (2) (2013) 164–168. [26] S.A. Owen, S.L. Hider, P. Martin, I.N. Bruce, A. Barton, W. Thomson, Genetic polymorphisms in key methotrexate pathway genes are associated with response to treatment in rheumatoid arthritis patients, Pharmacogenomics J. 13 (3) (2013) 227– 34. d. [27] T.N. Kung, J. Dennis, Y. Ma, G. Xie, V. Bykerk, J. Pope, C. Thorne, E. Keystone, K.A. Siminovitch, F. Gagnon, RFC1 80GNA is a genetic determinant of methotrexate efficacy in rheumatoid arthritis: a human genome epidemiologic review and metaanalysis of observational studies, Arthritis Rheum. 66 (5) (2014) 1111–1120. [28] Y. Ghodke-Puranik, A.S. Puranik, P. Shintre, K. Joshi, B. Patwardhan, J. Lamba, T.B. Niewold, A. Chopra, Folate metabolic pathway single nucleotide polymorphisms: a predictive pharmacogenetic marker of methotrexate response in Indian (Asian) patients with rheumatoid arthritis, Pharmacogenomics 16 (18) (2015) 2019–2034. [29] N. Muralidharan, C.M. Mariaselvam, M. Cb, V.S. Negi, Reduced folate carrier-1 80G NA gene polymorphism is not associated with methotrexate treatment response in South Indian Tamils with rheumatoid arthritis, Clin Rheumatol. 35 (4) (2016) 879–885 d. [30] R. Cáliz, J. del Amo, A. Balsa, F. Blanco, L. Silva, R. Sanmarti, F.G. Martínez, M.D. Collado, C. Ramirez Mdel, D. Tejedor, M. Artieda, D. Pascual-Salcedo, N. Oreiro, J.L. Andreu, E. Graell, L. Simon, A. Martínez, J. Mulero, The C677T polymorphism in the MTHFR gene is associated with the toxicity of methotrexate in a Spanish rheumatoid arthritis population, Scand J Rheumatol. 41 (1) (2012) 10–4. [31] J.C. Plaza-Plaza, M. Aguilera, M. Cañadas-Garre, C. Chemello, A. González-Utrilla, M.J. Faus Dader, M.A. Calleja, Pharmacogenetic polymorphisms contributing to toxicity induced by methotrexate in the southern Spanish population with rheumatoid arthritis, OMICS 16 (11) (2012 Nov) 589–595. [32] A. Lima, M. Bernardes, R. Azevedo, J. Monteiro, H. Sousa, R. Medeiros, V. Seabra, SLC19A1, SLC46A1 and SLCO1B1 polymorphisms as predictors of methotrexate-related toxicity in Portuguese rheumatoid arthritis patients, Toxicol. Sci. 142 (1) (2014) 196–209. [33] J. Świerkot, R. Ślęzak, P. Karpiński, J. Pawłowska, L. Noga, J. Szechiński, P. Wiland, Associations between single-nucleotide polymorphisms of RFC-1, GGH, MTHFR, TYMS, and TCII genes and the efficacy and toxicity of methotrexate treatment in patients with rheumatoid arthritis, Pol. Arch. Med. Wewn. 125 (3) (2015) 152–161. [34] P. Moya, J. Salazar, M.J. Arranz, C. Díaz-Torné, E. del Río, J. Casademont, H. Corominas, M. Baiget, Methotrexate pharmacokinetic genetic variants are associated with outcome in rheumatoid arthritis patients, Pharmacogenomics 17 (1) (2016) 25–29. [35] A. Lima, M. Bernardes, H. Sousa, R. Azevedo, L. Costa, F. Ventura, V. Seabra, R. Medeiros, SLC19A1 80G allele as a biomarker of methotrexate-related gastrointestinal toxicity in Portuguese rheumatoid arthritis patients, Pharmacogenomics 15 (6) (2014) 807–820. [36] M. Egger, G. Davey Smith, M. Schneider, C. Minder, Bias in meta-analysis detected by a simple, graphical test, BMJ 315 (7109) (1997 Sep 13) 629–634. [37] K. Fukino, T. Kawashima, M. Suzuki, K. Ueno, Methylenetetrahydrofolate reductase and reduced folate carrier-1 genotypes and methotrexate serum concentrations in patients with rheumatoid arthritis, J. Toxicol. Sci. 32 (4) (2007) 449–452. [38] Y. Ghodke, A. Chopra, P. Shintre, A. Puranik, K. Joshi, B. Patwardhan, Profiling single nucleotide polymorphisms (SNPs) across intracellular folate metabolic pathway in healthy Indians, Indian J. Med. Res. 133 (2011) 274–279. [39] P. Ranganathan, R. Culverhouse, S. Marsh, R. Ahluwalia, W.D. Shannon, S. Eisen, H.L. McLeod, Single nucleotide polymorphism profiling across the methotrexate pathway in normal subjects and patients with rheumatoid arthritis, Pharmacogenomics 5 (5) (2004) 559–569. [40] S. Sharma, M. Das, A. Kumar, V. Marwaha, S. Shankar, R. Aneja, R. Grover, V. Arya, V. Dhir, R. Gupta, U. Kumar, R.C. Juyal, T. B K, Interaction of genes from influx-metabolism-efflux pathway and their influence on methotrexate efficacy in rheumatoid arthritis patients among Indians, Pharmacogenet. Genomics 18 (12) (2008) 1041–1049. [41] H. Hayashi, C. Fujimaki, T. Daimon, S. Tsuboi, T. Matsuyama, K. Itoh, 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 (3) (2009) 355–361. [42] T. Kato, A. Hamada, S. Mori, H. Saito, 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 (2) (2012) 192–199. [43] T. Dervieux, N. Greenstein, J. Kremer, Pharmacogenomic and metabolic biomarkers in the folate pathway and their association with methotrexate effects during dosage escalation in rheumatoid arthritis, Arthritis Rheum. 54 (10) (2006) 3095–3103.

X. Li et al. / International Immunopharmacology 38 (2016) 8–15 [44] J.A. Wessels, J.K. de Vries-Bouwstra, B.T. Heijmans, P.E. Slagboom, Y.P. Goekoopuiterman, C.F. Allaart, P.J. Kerstens, D. van Zeben, F.C. Breedveld, B.A. Dijkmans, T.W. Huizinga, H.J. Guchelaar, Efficacy and toxicity of methotrexate in early rheumatoid arthritis are associated with single-nucleotide polymorphisms in genes coding for folate pathway enzymes, Arthritis Rheum. 54 (4) (2006) 1087–1095. [45] S.A. Samara, Y.M. Irshaid, K.N. Mustafa, Association of MDR1 C3435T and RFC1 80A polymorphisms with methotrexate toxicity and response in Jordanian rheumatoid arthritis patients, Int. J. Clin. Pharmacol. Ther. 52 (9) (2014) 746–755.

15

[46] H. Hayashi, Personalized drug therapy-directed clinical pharmacology research based on genetic polymorphisms and pharmacokinetics analysis, Yakugaku Zasshi. 133 (10) (2013) 1079–1085. [47] H.R. He, P. Liu, G.H. He, W.H. Dong, M.Y. Wang, Y.L. Dong, J. Lu, Association between reduced folate carrier G80A polymorphism and methotrexate toxicity in childhood acute lymphoblastic leukemia: a meta-analysis, Leuk. Lymphoma 55 (12) (2014) 2793–2800.