Systematic Review Pharmacogenomics

For reprint orders, please contact: reprints@futuremedicine.com

Are gene polymorphisms related to

treatment outcomes of methotrexate
in patients with rheumatoid arthritis?
A systematic review and meta-analysis

Aim: Identifying the predictors of responsiveness and adverse events in methotrexate Yuehong Chen‡,1, Kun Zou‡,2,
(MTX) treated patients with rheumatoid arthritis (RA) has been the focus of most Jianhong Sun1, Yuan Yang1
concern, but still without consistent consensus. Methods: PubMed and OVID EMBASE & Gang Liu*,1
1
Department of Rheumatology &
were searched to collect relevant studies that addressed correlations between gene
Immunology, West China Hospital,
polymorphisms and efficacy and/or toxicity in MTX-treated RA patients. Allelic, Sichuan University, Chengdu, China
recessive, dominant and over-dominant model were applied. Results: A total of 2
Department of Medical Record &
68 studies were included. For associations with efficacy, AMPD1 34C>T polymorphism Statistics, Sichuan Academy of Medical
was related to responsiveness in dominant model (odds ratio [OR]: 1.77; 95% CI: Sciences & Sichuan Provincial People’s
Hospital, Affiliated Hospital of University
1.19–2.63) and over-dominant model (OR: 1.59; 95% CI: 1.04–2.45). ATIC T675C
of Electronic Science & Technology,
polymorphism had association with responsiveness in recessive model (OR: 2.54; 95% Chengdu, China
CI: 1.23–5.26). For associations with toxicity, polymorphisms in TYMS 1494 del6 and *Author for correspondence:
FPGS rs10106 were correlated to absenting overall adverse events in recessive model Tel.: +86 288 542 2393
(OR: 0.68; 95% CI: 0.49–0.95) and dominant model (OR: 0.54; 95% CI: 0.35–0.83) Fax: +86 288 542 2394
neoliugang@163.com
respectively while MTHFR C677T was associated with presenting overall adverse ‡
Authors contributed equally
events in allelic model (OR: 1.29; 95% CI: 1.02–1.63), recessive model (OR: 1.38;
95% CI: 1.00–1.89) and dominant model (OR: 1.41; 95% CI: 1.02–1.94). Conclusion:
Polymorphisms in AMPD1 34C>T and ATIC T675C predict responsiveness. The absence
of TYMS 1494 del6 and FPGS rs10106 and presence of MTHFR C677T predict adverse
events in RA patients treated with MTX. Moreover, variations of the associations were
found between Caucasians and non-Caucasians.

First draft submitted: 18 September 2016; Accepted for publication: 16 October 2016;
Published online: 19 December 2016

Keywords:  efficacy • meta-analysis • methotrexate • polymorphisms • rheumatoid arthritis

• toxicity

Background different regions and populations worldwide

Rheumatoid arthritis (RA) is an inflamma-
tory autoimmune disease, of which patho-
genesis is largely unclear. It is generally
accepted that its aetiology is speculated to
be associated with genetics (the estimated
heritability as high as 65% for seropositive
RA), environmental factors (e.g., smok-
ing, low socioeconomic status), infections
(e.g., Epstein–Barr virus to trigger RA) and
microbiome (e.g., gut microbiome) [1] . The
incidence of RA varies from 0.1 to 5.0% in
with 5–50 new cases per 100,000 popula-
tion annually, placing huge burden to both
­individual, family and society [2–4] .
Methotrexate (MTX) is the most widely
used anchor disease-modifying anti­rheumatic
drug, either as monotherapy or combination
with other agents in nearly 83% RA patients
due to favorable cost–effectiveness ratio and
effectiveness. According to American Col-
lege of Rheumatology 20 response criteria,
part of
the responsiveness is reported in approxi-

10.2217/pgs-2016-0158 © 2017 Future Medicine Ltd Pharmacogenomics (2017) 18(2), 175–195 ISSN 1462-2416 175
Systematic Review  Chen, Zou, Sun, Yang & Liu
mately 70% patients after two years of MTX mono-
therapy  [5,6] . Nevertheless, the application of MTX
in RA patients also brings a concern, toxicity. The
overall adverse event rate at week 12 was reported as
45% and the discontinuation rate due to adverse events
was reported as 16% in MTX monotherapy group [7] ,
which might be associated with patient related, disease
related, treatment related and gene related variables
(e.g., genetic polymorphisms) [8] .
The MTX concentration being of wide inter-patient
variability in erythrocytes is reported to have a causal
relation with clinical efficacy that can be predicted by
the concentration measurement in RA patients [9,10]
and the metabolism of MTX depends on the vari-
ous enzymes and transporters, so gene variations in
pharmaco­k inetic enzymes and transporters may influ-
ence the treatment outcomes, efficacy as well as toxic-
ity, by affecting MTX metabolic effect and modifying
MTX transport. Therefore, identifying the predic-
tors of treatment response and toxicity is of crucial
importance in the management of RA to achieve
­individualized treatment  [11,12] .
Evidence on the association between gene polymor-
phisms and MTX treatment outcomes in RA patients
has been accumulating, but is still inconclusive and
inconsistent  [13] . We performed a systematic review
and meta-analysis to identify potentially gene predic-
tors of efficacy and toxicity in RA patients treated with
MTX.
Methods
This meta-analysis was reported in accordance with
the Preferred Reporting Items for Systematic Reviews
and Meta-Analyses guidelines (PRISMA) [14] .
Study selection criteria
We included articles studied in the associations
between any gene polymorphism and efficacy and/or
toxicity of MTX in RA patients. A study was excluded
if: it had no useful data for efficacy and toxicity; was an
editorial, comment, erratum, narrative review, letter or
conference abstract; was not published in English; was
a duplicate publication.
Search strategy
We systematically searched the electronic databases of
PubMed and OVID EMBASE from its inception to
April 20, 2016, using both MeSH terms and keywords,
such as ‘rheumatoid arthritis’, ‘methotrexate’, ‘gene’,
‘mutation’, ‘polymorphism’ and ‘variant’ (see Appendix
– Search Strategy). References of included studies and
published systematic reviews and meta-analyses were
also manually checked to identify ­potentially e­ligible
studies.
176 Pharmacogenomics (2017) 18(2)
Study selection
Two reviewers (Chen and Sun) independently screened
the potential eligible studies through titles, abstracts
and full-texts, using predefined study screening forms
and according to the inclusion and exclusion criteria.
Disagreements were solved by consensus. Study selec-
tion was conducted in Microsoft Office Access 2013
(Microsoft, WA, USA).
Data extraction
Pairs of four authors (Chen, Zou, Sun and Yang) inde-
pendently collected data from included studies including:
name of first author, year of publication, country, gene
sites, ethnicity of study population, number of MTX-
treated RA patients, mean age, mean disease duration,
mean dose of MTX, mean duration of MTX therapy,
numbers of responders and nonresponders, numbers
of present and absent adverse events, and genotype fre-
quencies of efficacy and toxicity. If genotype frequencies
of efficacy and/or toxicity were not reported, odds ratios
(OR) with 95% CI were extracted. If adjusted odds ratio
(ORadj) or adjusted hazard ratio (HRadj) was reported,
ORadj or HRadj with 95% CI and adjusted covari-
ates were collected. We used Microsoft Office Excel
(­Microsoft) to manage data extraction.
Toxic indicators were collected and classified accord-
ing to the system, such as: the overall adverse events,
gastro­intestinal, hepatic, CNS, dermato­logic, respiratory
and hematologic adverse events. ­Discontinues due to
adverse events were also recorded.
Statistical analysis
Heterogeneity among studies was assessed using the
Cochrane Q test (with a significant level of 0.1) and
I2statistic (25, 50, 75% indicating low, moderate and
high heterogeneity, respectively) [15] . The effect size of
association between gene polymorphisms and efficacy
(defined as responders vs nonresponders) and/or toxic-
ity (defined as present adverse events vs absent adverse
events) was estimated by ORs and corresponding
95% CIs; p-value of statistical significance was set at
0.05. Meta-analysis using random-effect model (Man-
tel–Haenszel method) was conducted if two or more
studies assessed the same gene polymorphism of effi-
cacy or toxicity. Subgroup analysis according to ethnic-
ity (Caucasians and non-Caucasians) was performed.
Four comparison models were analyzed which were:
allelic model (a vs A), recessive model (aa vs Aa+AA),
dominant model (aa+Aa vs AA) and over-dominant
model (Aa vs AA+aa); ‘a’ refers to a mutation allele and
‘A’ refers to a wild allele. All analyses were conducted
in Review Manager Software 5.1 (Cochrane Collabora-
tion, Oxford, UK). Sensitivity analysis was performed
to test the robust of results by pooling the ORs directly
future science group

obtained from primary studies that didn’t report raw
data to ORs calculated from raw data. Publication bias
was assessed by funnel plot.
Results
Totally, 2235 citations were retrieved from elec-
tronic databases, and after excluding duplications,
screening titles and abstracts and reading full-texts,
67 studies [16–82] were included (Figure 1) . An addi-
tional study [83] was included through manual search
of reference lists. Of the 68 eligible studies, 51 stud-
ies focused on Caucasians, 16 on non-Caucasians and
one on both Caucasians and non-Caucasians. A total
of 42 gene sites were assessed on efficacy and/or tox-
icity in MTX-treated RA patients, and the most fre-
quently reported gene polymorphism was MTHFR
C677T in 27 studies, followed by MTHFR A1298C in
19 studies and SLC19A1/RFC-1 80G/A in 16 studies
(see Supplementary Table 1). The involved gene poly-
morphisms reported only in one study was listed in
Supplementary Table 2.
Results from META-analysis
Association between gene polymorphisms
& efficacy in MTX-treated RA patients
Regardless of ethnicity, AMPD1 34C>T polymorphism
was related to treatment response in dominant model
(OR: 1.77; 95% CI: 1.19–2.63) and over-dominant
model (OR: 1.59; 95% CI: 1.04–2.45). ATIC T675C
polymorphism was correlated to treatment response in
recessive model (OR: 2.54; 95% CI: 1.23–5.26; see
Table 1).
In Caucasians, ABCB1/MDR-1 C3435T polymor-
phism was associated with nonresponse in recessive
model (OR: 0.57; 95% CI: 0.36–0.90) while associ-
ated with response in over-dominant model (OR: 1.60;
95% CI: 1.05–2.42). AMPD1 34C>T polymorphism
was related to response in dominant model (OR: 1.82;
95% CI: 1.08–3.07). SLC19A1/RFC-1 80G/A was
associated with response in recessive model (OR: 1.91;
95% CI: 1.16–3.15) (Table 1) .
In non-Caucasians, TYMS 28 bp (2R/3R) was
related to nonresponse in recessive model (OR: 0.47;
95% CI: 0.26–0.86). MTHFR A1298C associated
with nonresponse in recessive model (OR: 0.42; 95%
CI: 0.19–0.94) but with response in over-dominant
model (OR: 2.48; 95% CI: 1.42–4.32) (Table 1) .
Polymorphisms in TYMS 1494 del6, ABCB1/MDR-1
C1236T, ATIC 347C>G, GGH 452C>T, HLA-G
->+14 bp, ITPA 94C/A, MTHFR C677T, MTHFD1
1958 G/A, MTRR 66 A>G, MTR A2756G and
SHMT1 1420C/T were not associated with treatment
responsiveness and/or nonresponsiveness regardless of
ethnicity group (Table 1) .
future science group
Efficacy & toxicity of methotrexate  Systematic Review
Association between gene polymorphisms
& toxicity in MTX-treated RA patients
For overall adverse events, polymorphisms in TYMS
1494 del6 and FPGS rs10106 were related to absent-
ing overall adverse events in recessive model (OR:
0.68; 95% CI: 0.49–0.95) and in dominant model
(OR: 0.54; 95% CI: 0.35–0.83), respectively, in all
patients. MTHFR C677T polymorphism had asso-
ciation with presenting overall adverse events in all
patients (allelic model OR: 1.29; 95% CI: 1.02–1.63;
recessive model OR: 1.38; 95% CI: 1.00–1.89; domi-
nant model OR: 1.41; 95% CI: 1.02–1.94) and in
non-Caucasians (recessive model OR: 1.75; 95%
CI: 1.13–2.70; dominant model OR: 1.92; 95% CI:
1.01–3.64). TYMS 28 bp (2R/3R) was correlated to
presenting overall adverse events (dominant model
OR: 1.56; 95% CI: 1.07–2.26; over-dominant model
OR: 1.35; 95% CI: 1.00–1.81) in Caucasians while
absent adverse events (allelic model OR: 0.69; 95%
CI: 0.52–0.91; dominant model OR: 0.56; 95%
CI: 0.31–0.98) in non-Caucasians. ABCB1/MDR-1
C3435T related to absenting overall adverse events
(over-dominant model OR: 0.48; 95% CI: 0.23–0.97)
in Caucasians.
MS A2756G, ABCB1/MDR-1 C1236T, AMPD1
34C>T, GGH 401C/T, GGH 452C>T, GGH -354 G>T,
ITPA 94C/A, MTHFR A1298C, MTHFD1 1958 G/A,
MTRR 66 A>G, MTR A2756G, SHMT1 1420C/T
and SLC19A1/RFC-1 80G/A polymorphisms were not
2235 studies retrieved from electronic databases
(OVID EMBASE n = 1856, PubMed MEDLINE n = 379)
306 duplications removed
1929 studies screened through titles and abstracts
1758 not related studies exluded
171 studies selected through full texts
104 studies excluded:
44 conference abstracts
41 no useful data for efficacy/safety
9 systematic reviews/meta-analyses
3 duplicate studies
2 detecting genotypic distribution
2 gene expressions and efficacy
2 narrative reviews
1 nonEnglish published study
67 studies were included plus additional one study
identified through manual reference retrieval
Figure 1. Study selection flowchart.
www.futuremedicine.com 177
178
Table 1. Pooled results of the association between gene polymorphisms and efficacy in methotrexate-treated rheumatoid arthritis patients.
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city  Study R/NR I2 OR Study R/NR I2 OR Study R/NR I2 OR Study R/NR I2 OR
(n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)
TYMS 28 bp Total 8 1278/­ 49 0.84 6 510/­ 64 0.73 6 510/­ 0 0.77 6 510/­ 58 1.06
(2R/3R) 1112 (0.64–1.10) 462 (0.41–1.32) 462 (0.53–1.11) 462 (0.66–1.71)
  CA 5 1058/­ 58 0.90 4 417/­ 74 0.98 4 417/­ 0 0.82 4 417/­ 69 0.86
710 (0.65–1.25) 278 (0.42– 2.30) 278 (0.54–1.25) 278 (0.45–1.63)
  NCA 3 220/­ 36 0.71 2 93/­ 0 0.47 2 93/­ 0 0.63 2 93/­ 0 1.55
402 (0.40–1.25) 184 (0.26–0.86); 184 (0.30–1.30) 184 (0.87–2.77)
*p = 0.01 
TYMS Total 7 964/­ 59 1.03 5 354/­ 53 1.04 5 354/­ 52 0.90 5 354/­ 25 0.81
1494 del6 898 (0.71–1.49) 363 (0.50–2.14) 363 (0.47–1.72) 363 (0.54–1.22)

Pharmacogenomics (2017) 18(2)

  CA 3 598/­ 70 0.88 2 188/­ 58 0.93 2 188/­ 78 0.93 2 188/­ 0 0.74
424 (0.49–1.60) 143 (0.10–8.50) 143 (0.24–3.57) 143 (0.46–1.20)
Systematic Review  Chen, Zou, Sun, Yang & Liu

  NCA 4 366/­ 41 1.18 3 166/­ 56 1.23 3 166/­ 22 0.98 3 166/­ 51 0.80

474 (0.75–1.87) 220 (0.54–2.80) 220 (0.43–2.24) 220 (0.40–1.60)
ABCB1/ Total 7 1288/­ 76 0.93 7 644/­ 76 0.89 7 644/­ 52 0.89 7 644/­ 35 0.96
MDR-1 1212 (0.64–1.36) 606 (0.50–1.60) 606 (0.56–1.42) 606 (0.71–1.31)
C3435T
  CA 2 390/­ 0 0.78 2 195/­ 0 0.57 2 195/­ 0 0.98 2 195/­ 0 1.60
342 (0.58–1.05) 171 (0.36–0.90); 171 (0.59–1.63) 171 (1.05–2.42);
p = 0.01 p = 0.03
  NCA 3 638/­ 0 1.01 3 319/­ 0 1.21 3 319/­ 0 0.75 3 319/­ 0 0.71
572 (0.78–1.31) 286 (0.82–1.78) 286 (0.45–1.23)  286 (0.49–1.01)
ABCB1/ Total/ 2 180/ 0 1.10 2 90/­ 0 1.18 2 90/­ 0 1.08 2 90/­ 17 0.96
MDR-1 NCA ­356 (0.73–1.67) 178 (0.63–2.20) 178 (0.54–2.19) 178 (0.47–1.94)
C1236T
AMPD1 Total 3 812/­ 0 1.41 3 406/­ 0 0.68 4 438/­ 10 1.77 3 406/­ 0 1.59
34C>T 526 (0.96–2.08) 263 (0.11–4.24) 436 (1.19–2.63); 263 (1.04–2.45);
p = 0.005 p = 0.03
  CA 2 396/­ 11 1.39 2 198/­ 0 0.68 3 230/­ 39 1.82 2 198/­ 35 1.58
390 (0.89–2.17) 195 (0.11–4.24) 368 (1.08–3.07); 195 (0.88–2.86)
p = 0.02
ATIC Total 7 1638/­ 1 0.91 7 819/­ 0 0.99 8 853/­ 22 0.85 7 819/­ 40 0.85
347C>G 1164 (0.77–1.08) 582 (0.73–1.34) 756 (0.64–1.13) 582 (0.62–1.16)
  CA 3 534/­ 0 0.91 3 282/­ 0 0.76 4 316/­ 62 0.85 3 282/­ 73 0.87
250 (0.66–1.27) 210 (0.42–1.38) 384 (0.47–1.54) 210 (0.39–1.93)
*Boldface is for statistically significant results.

future science group

CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; R/NR: Responder/nonresponder.
 Table 1. Pooled results of the association between gene polymorphisms and efficacy in methotrexate-treated rheumatoid arthritis patients (cont.).
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city  Study R/NR I2
OR Study R/NR 2
I OR Study R/NR I 2
OR Study R/NR I2 OR
(n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)
ATIC NCA 4 1104/­ 34 0.91 4 537/­ 0 1.08 4 537/­ 0 0.86 4 537/­ 0 0.85

future science group

347C>G 914 (0.71–1.16) 372 (0.76–1.55) 372 (0.60–1.24) 372 (0.63–1.14)
(cont.)
ATIC T675C Total/ 2 272/­ 38 1.62 2 136/­ 0 2.54 2 136/­ 80 1.94 2 136/­ 86 1.35
CA 426 (1.05–2.52); 213 (1.23–5.26); 213 (0.58–6.52) 213 (0.36–5.02)
p = 0.03 p = 0.01
GGH Total/ 3 678/­ 2 1.18 3 339/­ NA 2.18 3 339/­ 3 1.17 3 339/­ 0 1.14
452C>T CA 426 (0.72–1.93) 213 (0.09–54.13) 213 (0.70–1.98) 213 (0.68–1.90)
HLA-G Total/ 3 498/­ 48 0.94 3 249/­ 0 0.90 3 249/­ 80 1.00 3 249/­ 84 1.04
->+14 bp CA 404 (0.65–1.36) 202 (0.57–1.44) 202 (0.40–2.49) 202 (0.40–2.71)
ITPA 94C/A Total/ 3 462/­ 40 0.63 3 231/­ NA 0.82 4 265/­ 47 0.53 3 231/­ 41 0.63
CA 560 (0.30–1.30) 280 (0.03–20.54) 457 (0.23–1.21) 280 (0.29–1.35)
MTHFR Total 12 1656/­ 10 0.97 11 811/­ 0 0.91 15 1142/­ 0 0.92 11 811/­ 0 1.00
C677T 1700 (0.82–1.15) 833 (0.65–1.28) 1113 (0.77–1.10) 833 (0.81–1.23)
  CA 9 1400/­ 19 0.94 9 700/­ 0 0.88 10 862/­ 0 0.89 9 700/­ 0 0.99
1268 (0.77–1.13) 634 (0.62–1.25) 878 (0.73–1.09) 634 (0.79–1.25)
  NCA 3 256/­ 0 1.21 2 111/­ 0 1.32 5 280/­ 0 1.02 2 111/­ 0 1.03
432 (0.78–1.86) 199 (0.45–3.86) 235 (0.70–1.49) 199 (0.57–1.84)
MTHFR Total 11 1290/­ 17 0.96 10 628/­ 30 0.89 13 878/­ 52 1.05 10 628/­ 51 1.13
A1298C 1436 (0.78–1.16) 701 (0.54–1.48) 948 (0.77–1.43) 701 (0.79–1.60)
  CA 8 1034/­ 0 0.96 8 517/­ 11 1.09 9 658/­ 0 0.93 8 517/­ 3 0.89
1004 (0.79–1.17) 502 (0.67–1.78) 600 (0.73–1.18) 502 (0.68–1.16)
  NCA 3 256/­ 69 1.09 2 111/­ 0 0.42 4 220/­ 83 1.42 2 111/­ 0 2.48
432 (0.48–2.48) 199 (0.19–0.94); 348 (0.51–3.97) 199 (1.42–4.32);
p = 0.04 p = 0.001
MTHFD1 Total/ 2 396/­ 69 0.84 2 198/­ 15 0.66 3 232/­ 68 0.73 2 198/­ 0 1.22
1958 G/A CA 350 (0.50–1.42) 175 (0.38–1.16) 349 (0.36–1.45) 175 (0.81–1.84)
Efficacy & toxicity of methotrexate 

MTRR 66 Total 2 316/­ 0 1.09 2 158/­ 0 1.08 2 158/­ 0 1.17 2 158/­ 0 1.06
A>G 488 (0.81–1.49) 244 (0.67–1.73) 244 (0.72–1.90) 244 (0.69–1.64)

www.futuremedicine.com
MTR Total/ 2 390/­ 78 0.71 2 195/­ 89 0.30 2 195/­ 0 1.05 2 195/­ 67 2.41
A2756G CA 184 (0.24–2.06) 92 (0.01–9.38) 92 (0.62–1.80) 92 (0.21–27.89)
SHMT1 Total 4 658/­ 0 1.01 4 329/­ 0 1.24 4 329/­ 0 0.89 4 329/­ 0 0.81
1420C/T 718 (0.77–1.33) 359 (0.75–2.08) 359 (0.61–1.30) 359 (0.57–1.15)
*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; R/NR: Responder/nonresponder.
Systematic Review

179
Systematic Review  Chen, Zou, Sun, Yang & Liu

related to presenting and/or absenting overall adverse

(0.53–1.35)

(0.77–1.60)
(0.74–1.27)
(0.51–1.11)
events regardless of ethnicity group (Table 2) .
Table 1. Pooled results of the association between gene polymorphisms and efficacy in methotrexate-treated rheumatoid arthritis patients (cont.).

(95% CI)

For gastrointestinal adverse events, only ATIC 347C>G

0.85
0.97
0.75

polymorphism was associated with presenting adverse

1.11
OR
Over-dominant model

events in overall patients in recessive model (OR: 1.76;

95% CI: 1.03–3.03) while other gene polymorphisms
(%)

21
I2

0
were not related to presenting and/or absenting adverse
events regardless of ethnicity group. For dermato­logic
Study R/NR

280/­

240/­
492/­

252/­

307
207
191

514

adverse events, RFC1 80 G/A polymorphism was associ-

ated with absenting adverse events in dominant model
(OR: 0.41; 95% CI: 0.21–0.80) in overall patients and
(n)
3

3
Caucasians while other gene polymorphisms were not
correlated to presenting and/or absenting adverse events
(0.66–4.54)

(0.58–1.49)
(0.73–1.94)
(0.61–1.33)

regardless of ethnicity group (Table 2).

(%) (95% CI)

No associations between genetic variations and adverse

Dominant model

0.90

0.93
1.74
1.19
OR

events in hepatic, central nervous system, respiratory and

hematologic system, and discontinuation due to adverse
492/­ 56

240/­ 14
252/­ 74

events were found in all genetic models (Table 2).

Study R/NR I2

280/­ 0

307
207
191

514

Data not included in meta-analysis

Unadjusted data of association between gene
polymorphisms & MTX treatment outcome in RA
(n)
3

patients
Two primary studies that data were not eligible for
(0.86–3.78)

(0.70–2.25)

(0.38–1.89)
(1.16–3.15);

CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; R/NR: Responder/nonresponder.

pooled analysis reported unadjusted ORs and 95%

(95% CI)

p = 0.01

CIs between gene polymorphisms and efficacy and/or

0.85
1.80

1.25

1.91
Recessive model

adverse events in MTX treated RA patients [32,83] .

OR

Morgan [32] reported MTHFR C677T and MTHFR

(%)

A1298C polymorphisms were not related to treatment

63

67
I2

response in homozygous model (OR: 0.67; 95% CI:

Study R/NR

280/­

240/­
492/­

252/­

0.30–1.49; OR: 0.99; 95% CI: 0.45–2.17) and hetero­

307
207
191

514

zygous model (OR: 0.76; 95% CI: 0.45–1.26; OR:

1.34; 95% CI: 0.81–2.22) (Supplementary Table 3) .
Weisman et al. [83] reported ATIC C347G was related
(n)
3

to increased overall adverse events (OR: 2.42; 95% CI:

(0.88–2.91)

1.07–5.49) and gastrointestinal adverse events (OR:

(0.55–1.42)
(0.79–1.70)
(0.77–1.42)
(%) (95% CI)

2.97; 95% CI: 1.64–5.36) in recessive model; MTHFR

C677T was associated with elevated adverse events of
0.88
1.60
1.05

1.16
OR

central nervous system in recessive model (OR: 3.33;

Allelic model

95% CI: 1.72–6.45); and TSER*2/*3 was associ-

68
72

72
I2

ated with decreased overall adverse events (OR: 0.54;

95% CI: 0.35–0.85) and alopecia (OR: 0.19; 95% CI:
Study R/NR

480/­
984/­

504/­
560/­

1028

0.11–0.33) in dominant model (Supplementary Table 3).

382

414

614
*Boldface is for statistically significant results.

Adjusted data of association between gene

(n)

polymorphisms & MTX treatment outcome in RA

3

patients
Due to limited data, we were unable to pool adjusted
Ethni­

Total

NCA
city 

ORs of associations between gene polymorphisms and

CA

CA

efficacy and toxicity in MTX-treated RA patients. Age

and gender were the most commonly adjusted factors
SLC19A1/
1420C/T
Genetic

(Supplementary Table 3) .
SHMT1

(cont.)

80G/A
RFC-1
sites 

A total of 14 studies reported adjusted ORs of the

associations between gene polymorphisms and effi-
 

180 Pharmacogenomics (2017) 18(2) future science group

 Table 2. Pooled results of the association between gene polymorphisms and adverse events in methotrexate treated rheumatoid arthritis patients
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR
  (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)
Overall adverse events 

future science group

TYMS 2 Total 7 1090/­ 56 0.98 6 532/­ 18 0.87 6 532/­ 56 1.17 6 532/­ 0 1.24
8 bp 1414 (0.75–1.28) 686 (0.65–1.16) 686 (0.71–1.93) 686 (0.98–1.57)
(2R/3R)
  CA 4 630/­ 25 1.15 4 315/­ 7 0.99 4 315/­ 0 1.56 4 315/­ 0 1.35
948 (0.90–1.47) 474 (0.70–1.40) 474 (1.07–2.26); 474 (1.00–1.81);
*p = 0.02 p = 0.05
  NCA 3 460/­ 0 0.69 2 217/­ 18 0.71 2 217/­ 0 0.56 2 217/­ 0 1.07
466 (0.52–0.91); 212 (0.45–1.13) 212 (0.31–0.98); 212 (0.73–1.59)
p = 0.009 p = 0.04
TYMS Total 6 972/­ 49 0.89 5 473/­ 0 0.68 5 473/­ 67 0.88 5 473/­ 2 1.15
1494 del6 1156 (0.68–1.18) 557 (0.49–0.95); 557 (0.49–1.59) 557 (0.89–1.48)
p = 0.02
  CA 2 406/­ 0 1.03 2 203/­ 64 0.60 2 203/­ 0 1.20 2 203/­ 0 1.40
532 (0.77–1.38) 266 (0.16–2.30) 266 (0.83–1.75) 266 (0.96–2.05)
  NCA 4 566/­ 48 0.82 3 270/­ 0 0.69 3 270/­ 54 0.65 3 270/­ 0 0.98
624 (0.57–1.20) 291 (0.48–0.98); 291 (0.25–1.65) 291 (0.70–1.37)
p = 0.04
MS Total – – – – – – – – 2 283/­ 70 0.98 – – – –
A2756G 188 (0.42–2.29)
ABCB1/ Total 5 896/­ 57 0.86 5 448/­ 60 0.94 7 665/­ 63 0.79 5 448/­ 73 0.81
MDR-1 1150 (0.63–1.17) 575 (0.56–1.58) 628 (0.45–1.38) 575 (0.46–1.41)
C3435T CA 2 326/­ 55 0.64 2 163/­ 0 0.99 4 380/­ 77 0.56 2 163/­78 27 0.48
  156 (0.33–1.24) 78 (0.53–1.83) 131 (0.17–1.83) (0.23–0.97);
p = 0.04
  NCA 2 474/­ 76 0.90 2 237/­ 85 0.71 2 237/­ 0 1.12 2 237/­ 75 1.37
842 (0.54–1.49) 421 (0.25–2.00) 421 (0.70–1.79) 421 (0.68–2.79)
ABCB1/ Total 2 356/­ 76 2.65 2 178/­ NA 1.14 2 178/­ 83 4.04 2 178/­ 84 3.82
Efficacy & toxicity of methotrexate 

MDR-1 322 (0.27–25.69) 161 (0.71–1.81) 161 (0.16–99.28) 161 (0.14–

C1236T 105.12)

www.futuremedicine.com
FPGS Total/ – – – – – – – – 2 125/­ 0 0.54 – – – –
rs10106 CA 319 (0.35–0.83);
p = 0.005
*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; P/A: Patient number of present adverse event/absent adverse event.

181
Systematic Review
182
Table 2. Pooled results of the association between gene polymorphisms and adverse events in methotrexate treated rheumatoid arthritis patients
(cont.).
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR
  (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)
AMPD1 Total/ 2 394/­ 0 1.20 2 197/­ 0 0.58 3 339/­ 17 1.04 2 197/­ 0 1.38
34C>T CA 384 (0.75–1.91) 192 (0.10–3.40) 255 (0.66–1.65) 192 (0.81–2.35)
ATIC Total 4 824/­ 68 1.12 4 412/­ 55 1.12 5 556/­ 38 1.15 4 412/­ 33 1.21
347C>G 830 (0.75–1.66) 415 (0.54–2.32) 479 (0.80–1.64) 415 (0.84–1.74)
  CA 2 396/­ 0 1.38 2 198/­ 0 1.89 3 342/­ 34 1.30 2 198/­ 76 1.09
382 (0.99–1.91) 191 (0.87–4.11) 255 (0.83–2.02) 191 (0.44–2.71)
  NCA 2 428/­ 65 0.91 2 214/­ 0 0.63 2 214/­ 50 0.97 2 214/­ 0 1.29
448 (0.52–1.61) 224 (0.39–1.00); 224 (0.51–1.84) 224 (0.88–1.89)
p = 0.05

Pharmacogenomics (2017) 18(2)

GGH Total 2 596/­ 76 1.17 2 298/­ 45 1.34 2 298/­ 83 0.78 2 298/­ 0 0.80
401C/T 526 (0.70–1.97) 263 (0.85–2.12) 263 (0.14–4.19) 263 (0.57–1.12)
Systematic Review  Chen, Zou, Sun, Yang & Liu

GGH Total/ 2 292/­ 65 0.86 2 146/­ 0 0.60 2 146/­ 74 0.93 2 146/­ 71 1.08
rs7010484 CA 580 (0.50–1.46) 290 (0.28–1.30) 290 (0.41–2.08) 290 (0.50–2.31)
T/C
GGH Total/ 3 540/­ 0 1.32 3 270/­ 0 1.90 3 270/­ 0 1.32 3 270/­ 0 1.21
452C>T CA 580 (0.92–1.89) 290 (0.57–6.35) 290 (0.87–2.00) 290 (0.79–1.84)
GGH Total/ 2 236/­ 0 0.76 2 118/­ 0 0.53 2 118/­ 0 0.75 2 118/­ 0 0.87
-354 G>T CA 634 (0.53–1.07) 317 (0.20–1.45) 317 (0.49–1.15) 317 (0.56–1.34)
ITPA Total/ 2 392/­ 0 0.79 2 196/­ NA Not 3 342/­ 20 1.02 2 196/­ 0 0.78
94C/A CA 382 (0.43–1.45) 191 estimable 256 (0.56–1.87) 191 (0.41–1.46)
MTHFR Total 16 2312/­ 66 1.29 15 1143/­ 23 1.38 20 1409/­ 72 1.41 15 1143/­ 52 1.20
C677T 3076 (1.02–1.63); 1517 (1.00–1.89); 2049 (1.02–1.94); 1517 (0.92–1.56)
p = 0.03 p = 0.05 p = 0.05
  CA 11 1440/­ 60 1.24 11 720/­ 32 1.24 13 892/­ 59 1.17 11 720/­ 41 1.12
2046 (0.94–1.63) 1023 (0.81–1.90) 1241 (0.84–1.63) 1023 (0.83–1.51)
  NCA 5 872/­ 73 1.39 4 423/­ 0 1.75 7 517/­ 80 1.92 4 423/­ 67 1.33
1030 (0.87–2.19) 494 (1.13–2.70); 808 (1.01–3.64); 494 (0.79–2.26)
p = 0.01 p = 0.05
MTHFR Total 13 1748/­ 65 0.91 13 974/­ 40 0.96 15 985/­ 44 0.98 12 861/­ 21 0.94
A1298C 2538 (0.70–1.18) 1285 (0.61–1.51) 1621 (0.76–1.25) 1248 (0.75–1.18)
  CA 9 1186/­ 73 0.94 10 706/­ 52 0.94 10 653/­ 63 0.99 9 593/­ 39 0.96
1970 (0.66–1.34) 1022 (0.51–1.75) 1164 (0.69–1.42) 985 (0.70–1.31)
*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; P/A: Patient number of present adverse event/absent adverse event.

future science group

 Table 2. Pooled results of the association between gene polymorphisms and adverse events in methotrexate treated rheumatoid arthritis patients
(cont.).
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city Study P/A 2
I OR Study P/A 2
I OR Study P/A 2
I OR Study P/A I2 OR
  (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)

future science group

MTHFR NCA 4 562/­ 31 0.85 3 268/­ 0 1.08 5 332/­ 0 0.94 3 268/­ 0 0.89
A1298C 568 (0.59–1.22) 263 (0.68–1.73) 457 (0.68–1.29) 263 (0.62–1.27)
(cont.)
MTHFD1 Total/ – – – – – – – – 2 280/­ 70 0.86 – – – –
1958 G/A CA 115 (0.34–2.15)
MTRR 66 Total 3 728/­ 51 0.83 3 364/­ 62 0.91 4 477/­ 0 0.71 3 364/­ 39 0.83
A>G 568 (0.59–1.17) 284 (0.50–1.66) 321 (0.50–1.01) 284 (0.53–1.29)
  CA 2 388/­ 72 0.77 2 194/­ 75 0.79 3 307/­ 0 0.74 2 194/­ 57 0.99
264 (0.41–1.43) 132 (0.29–2.17) 169 (0.46–1.18) 132 (0.46–2.12)
MTR Total/ 3 456/­ 10 1.06 3 228/­ 25 1.29 3 228/­ 9 1.00 3 228/­ 23 0.90
A2756G CA 376 (0.73–1.53) 188 (0.53–3.12) 188 (0.63–1.59) 188 (0.53–1.52)
SHMT1 Total 2 600/­ 42 0.81 2 300/­ 0 0.68 2 300/­ 43 0.83 2 300/­ 0 1.37
1420C/T 398 (0.52–1.25) 199 (0.44–1.06) 199 (0.29–2.37) 199 (0.91–2.08)
SLC19A1/ Total 9 1488/­ 0 0.89 9 744/­ 0 0.86 11 961/­ 9 0.85 9 744/­ 29 0.98
RFC-1 2214 (0.77–1.03) 1107 (0.66–1.12) 1160 (0.68–1.06) 1107 (0.76–1.26)
80G/A CA 6 918/­ 0 0.92 6 459/­ 13 0.87 8 676/­ 23 0.89 6 459/­ 42 1.02
  1238 (0.76–1.11) 619 (0.60–1.27) 672 (0.64–1.22) 619 (0.70–1.49)
  NCA 3 570/­ 0 0.85 3 285/­ 2 0.85 3 285/­ 0 0.78 3 285/­ 4 0.89
976 (0.68–1.07) 488 (0.55–1.30) 488 (0.56–1.08) 488 (0.65–1.22)
Gastrointestinal adverse events
ATIC Total 2 230/­ 30 1.43 2 115/­ 0 1.76 2 115/­ 61 2.04 2 115/­ 0 0.93
347C>G 786 (0.96– 2.13) 393 (1.03–3.03); 393 (0.42–9.92) 393 (0.59–1.45)
p = 0.04
ABCC2 Total 2 166/­ 42 1.73 2 83/­ 60 7.07 2 83/­ 0 1.26 2 83/­ 37 0.75
1249 G/A 264 (0.65–4.60) 132 (0.38–131.06) 132 (0.69–2.31) 132 (0.15–3.75)
MTR Total/ 2 180/­ 0 1.02 2 90/­ 67 0.87 2 90/­ 0 1.10 2 90/­ 18 1.15
Efficacy & toxicity of methotrexate 

A2756G CA 370 (0.65–1.58) 185 (0.13–5.65) 185 (0.64–1.90) 185 (0.58–2.28)

SLC19A1/ Total/ 2 272/­ 62 0.78 2 136/­ 68 0.59 3 189/­ 38 0.73 2 136/­ 0 1.36

www.futuremedicine.com
RFC1 CA 568 (0.48–1.28) 284 (0.24–1.45) 351 (0.43–1.23) 284 (0.89–2.06)
G80A
MTHFR Total 3 334/­ 23 0.91 3 167/­ 50 1.15 5 235/­ 56 0.93 3 167/­ 47 0.76
C677T 556 (0.65–1.27) 278 (0.53–2.47) 745 (0.53–1.63) 278 (0.42–1.36)
*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; P/A: Patient number of present adverse event/absent adverse event.

183
Systematic Review
184
Table 2. Pooled results of the association between gene polymorphisms and adverse events in methotrexate treated rheumatoid arthritis patients
(cont.)
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR
  (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)
  CA 2 186/­ 0 1.10 2 93/­ 59 1.67 2 93/­ 0 1.02 2 93/­ 72 0.67
370 (0.76–1.60) 185 (0.49–5.62) 185 (0.61–1.72) 185 (0.19–2.35)
non-  –  –  –  –  –  –    – 3 142/­ 76 0.88  –  – –  – 
CA 560 (0.31–2.50)
MTHFR Total 2 304/­ 70 0.82 2 152/­ 0 1.14 2 152/­ 62 0.77 2 152/­ 37 0.76
A1298C 404 (0.41–1.61) 202 (0.43–3.05) 202 (0.37–1.61) 202 (0.43–1.35)
SHMT1 Total – – – – – – – – 2 97/­ 87 2.91 – – – –
C1420T 402 (0.60–14.18)

Pharmacogenomics (2017) 18(2)

ABCB1/ Total/ – – – – – – – – 2 132/­ 0 0.78 – – – –
MDR1 CA 177 (0.46–1.32)
Systematic Review  Chen, Zou, Sun, Yang & Liu

C3435T
Hepatic adverse events
MTHFR Total 3 182/­ 5 1.20 3 91/­ 0 1.05 5 150/­ 64 1.58 3 91/­ 39 1.39
C677T 772 (0.84–1.71) 386 (0.55–2.01) 862 (0.73–3.43) 386 (0.70–2.74)
  CA 2 62/­ 36 1.35 2 31/­ 0 0.97 2 31/­ 58 1.81 2 31/­ 48 1.80
558 (0.68–2.68) 279 (0.32–2.98) 279 (0.47–6.90) 279 (0.60–5.35)
  NCA – – – – –   – – 3 119/­ 77 1.43 – – – –
583 (0.45–4.48)
MTHFR Total 2 146/­ 81 1.39 2 73/­ 48 1.81 2 73/­ 77 1.48 2 73/­ 50 1.15
A1298C 328 (0.40–4.87) 164 (0.22–15.19) 164 (0.32–6.80) 164 (0.44–3.01)
RFC1 80 Total/ – – – – – – – – 2 48/­ 0 0.78 – – – –
G/A CA 331 (0.40–1.51)
CNS adverse events
SLC19A1/ Total/ – – – – – – – – 2 130/­ 54 0.96 – – – –
RFC1 CA 177 (0.38–2.46)
80G/A
ABCB1/ Total/ – – – – – – – – 2 132/­ 54 0.69 – – – –
MDR-1 CA 177 (0.25–1.87)
C3435T
MTHFR Total 2 282/­ 21 0.91 2 141/­ 0 0.81 3 143/­ 3 0.94 2 141/­ 0 1.10
C677T 426 (0.61–1.36) 213 (0.42–1.54) 596 (0.57–1.54) 213 (0.68–1.78)
  NCA – – – – – – – – 2 30/­ 0 0.58 – – – –
522 (0.25–1.32)

future science group

*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; P/A: Patient number of present adverse event/absent adverse event.
 Table 2. Pooled results of the association between gene polymorphisms and adverse events in methotrexate treated rheumatoid arthritis patients
(cont.)
Genetic Ethni­ Allelic model Recessive model Dominant model Over-dominant model
sites  city Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR Study P/A I2 OR
  (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI) (n) (%) (95% CI)

future science group

MTHFR Total 2 282/­ 0 1.00 2 141/­ 0 0.62 2 141/­ 0 1.15 2 141/­ 0 1.31
A1298C 426 (0.68–1.48) 213 (0.22–1.72) 213 (0.70–1.87) 213 (0.80–2.14)
Dermatologic adverse events
RFC1 Total/ – – – – – – – – 2 46/­ 0 0.41 – – – –
80 G/A CA 341 (0.21–0.80);
p = 0.009
MTHFR Total/ 2 22/­ 80 3.01 2 11/­ NA 0.59 3 64/­ 50 3.30 2 11/­189 71 4.92
C677T CA 378 (0.20–45.95) 189 (0.07–4.97) 521 (0.92–11.89) (0.27–89.17)
Respiratory adverse events
MTHFR Total/ – – – – – – – – 2 4/­531 0 0.93 – – – –
C677T NCA (0.13–6.44)
Hematologic adverse events
MTHFR Total/ – – – – – – – – 3 13/­ 0 1.13 – – – –
C677T NCA 689 (0.37–3.46)
Discontinuation due to adverse events
MTHFR Total – – – – – – – – 2 66/­ 77 0.86 – – – –
C677T 439 (0.23–3.25)
*Boldface is for statistically significant results.
CA: Caucasian; NA: Not applicable; NCA: Non-Caucasian; no.: Number; OR: Odds ratio; P/A: Patient number of present adverse event/absent adverse event.
Efficacy & toxicity of methotrexate 

www.futuremedicine.com
185
Systematic Review
186
Table 3. Associations between genetic polymorphisms and efficacy and/or toxicity of methotrexate in patients with rheumatoid arthritis reported by
previously published meta-analyses.
Study (year)  Genetic Outcomes Genetic model Overall patients Caucasians Asians Association with Ref.
polymorphism  (OR [95% CI], (OR [95% CI], (OR [95% CI], methotrexate
    study number)  study number) study number)  Efficacy Toxicity  
Lee et al. ATIC 347 C/G Non­ Allelic model 1.19 (0.96–1.48), 1.74 (1.18–2.56), 1.01 (0.78–1.31), Maybe Maybe [87]
(2016) responsiveness n=4 n=2 n=2
      Recessive model 1.53 (1.13–2.06), 2.19 (1.31–3.65), 1.27 (0.88–1.83),      
n=4 n=2 n=2
      Dominant model 1.57 (1.15–2.16), 1.88 (1.24–2.87), 1.25 (0.77–2.01),      
n=5 n=3 n=2
      Over-dominant 1.53 (1.13–2.06), 2.19 (1.31–3.65), 1.27 (0.88–1.83),      
model n=4 n=2 n=2

Pharmacogenomics (2017) 18(2)

    Toxicity Allelic model 1.30 (1.00–1.70), 1.61 (1.02–2.53), 1.63 (0.84–1.62),      
n=3 n=1 n=2
Systematic Review  Chen, Zou, Sun, Yang & Liu

      Recessive model 1.19 (0.73–1.96), 1.61 (0.55–4.77), 1.10 (0.63–1.93),      

n=3 n=1 n=2
      Dominant model 1.45 (1.03–2.04), 1.74 (1.08–2.81), 1.21 (0.74–1.96),      
n=4 n=2 n=2
      Over-dominant 1.34 (0.94–1.93), 1.73 (0.94–3.18), 1.17 (0.75–1.84),      
model n=3 n=1 n=2
Li et al. (2016) RFC1 80G/A Efficacy Allelic model 1.29 (1.05–1.67), 1.24 (0.93–1.67), 1.42 (1.14–1.77), Yes Maybe [88]
n = 12 n=8 n=4 not
      Recessive model 1.49 (1.17–1.90), 1.37 (1.00–1.86), 1.71 (1.15–2.55),      
n = 12 n=8 n=4
      Dominant model 0.73 (0.53–1.01), 0.78 (0.49–1.23), 0.63 (0.45–0.90),      
n = 12 n=8 n=4
      Over-dominant 0.97 (0.80–1.17), 0.93 (0.73–1.19), 1.02 (0.75–1.40),      
model n = 12 n=8 n=4
    Toxicity Allelic model 0.92 (0.81–1.04), 0.87 (0.69–1.10), 1.05 (0.84–1.31),      
n = 12 n=9 n=3
      Recessive model 0.91 (0.74–1.12), 0.90 (0.69–1.17), 0.92 (0.65–1.32),      
n = 12 n=9 n=3
      Dominant model 1.16 (0.95–1.41), 1.11 (0.88–1.40), 1.30 (0.90–1.89),      
n = 12 n=9 n=3
      Over-dominant 0.95 (0.80–1.14), 0.98 (0.78–1.24), 0.89 (0.65–1.20),      
model n = 12 n=9 n=3
EA: East Asian; NA: Not acquired; NC: Non-Caucasian; NOA: Non-Asian.

future science group

 Table 3. Associations between genetic polymorphisms and efficacy and/or toxicity of methotrexate in patients with rheumatoid arthritis reported by
previously published meta-analyses (cont.).
Study (year)  Genetic Outcomes Genetic model Overall patients Caucasians Asians Association with Ref.
polymorphism  (OR [95% CI], (OR [95% CI], (OR [95% CI], methotrexate
    study number)  study number) study number)  Efficacy Toxicity  

future science group

Lee et al. ABCB1 C3435T Non­ Allelic model 1.20 (0.46–3.16), NA NA Maybe Maybe [89]
(2016) responsiveness n=3 Not
      Recessive model 1.47 (0.35–6.15), NA NA      
n=3
      Dominant model 1.08 (0.34–3.44), NA NA      
n=3
      Over-dominant 0.94 (0.64–1.38), NA NA      
model n=3
    Toxicity Allelic model 1.09 (0.44–2.70), 1.03 (0.17–6.20), NA      
n=3 n=2
      Recessive model 1.62 (0.88–2.97), 1.39 (0.62–3.12), NA      
n=3 n=2
      Dominant model 0.79 (0.18–3.40), 0.14 (0.03–0.57), NA      
n=3 n=1
      Over-dominant 0.48 (0.26–0.90), 0.28 (0.09–0.87), NA      
model n=2 n=1
Kung et al. RFC1 80G>A Efficacy Allelic model 1.28 (1.10–1.49), 1.24 (1.05–1.47), NA Yes No [90]
(2014) n = 10 n=8
      Recessive model 1.42 (1.04–1.93), 1.34 (0.98–1.83), NA      
n = 10 n=8
      Dominant model 1.32 (0.89–1.94), NA NA      
n = 10
      Over-dominant 0.93 (0.75–1.15), NA NA      
model n = 10
    Toxicity Allelic model 0.97 (0.80–1.17), 0.98 (0.79–1.21), NA      
n=7 n=6
Efficacy & toxicity of methotrexate 

      Recessive model 0.93 (0.65–1.33), 0.98 (0.66–1.47), NA      

n=7 n=6

www.futuremedicine.com
      Dominant model 0.97 (0.70–1.33), NA NA      
n=7
      Over-dominant 0.98 (0.75–1.27), NA NA      
model n=7
EA: East Asian; NA: Not acquired; NC: Non-Caucasian; NOA: Non-Asian.

187
Systematic Review
188
Table 3. Associations between genetic polymorphisms and efficacy and/or toxicity of methotrexate in patients with rheumatoid arthritis reported by
previously published meta-analyses (cont.).
Study (year)  Genetic Outcomes Genetic model Overall patients Caucasians Asians Association with Ref.
polymorphism  (OR [95% CI], (OR [95% CI], (OR [95% CI], methotrexate
    study number)  study number) study number)  Efficacy Toxicity  
Morgan et al. MTHFR C677T Efficacy CT vs CC 0.82 (0.60–1.11), NA NA Unlikely NA [32]
(2014) n=4 to have
      TT vs CC 0.76 (0.46–1.24), NA NA      
n=4
  MTHFR A1298C Efficacy AA vs AC 0.91 (0.65–1.25), NA NA Unlikely NA  
n=3 to have
      CC vs AA 1.09 (0.64–1.86), NA NA      
n=3

Pharmacogenomics (2017) 18(2)

Song et al. MTHFR C677T Toxicity Recessive model 1.62 (1.19–2.20), 1.95 (1.08–3.53), 1.58 (1.08–2.33), NA Yes [91]
(2014) n=7 n=1 n = 4 (EA)
Systematic Review  Chen, Zou, Sun, Yang & Liu

      Dominant model 1.46 (0.92–2.30), 1.21 (0.57–2.56), 2.07 (0.95–4.52),      

n = 11 n=3 n = 5 (EA)
      TT vs CC 1.95 (1.05–3.60), 2.13 (1.09–4.15), 2.27 (0.82–6.33),      
n=7 n=1 n = 4 (EA)
  MTHFR A1298C Toxicity Recessive model 0.76 (0.39–1.45), 0.98 (0.58–1.66), 0.54 (0.15–1.93), NA Yes  
n=8 n=4 n = 3 (EA)
      Dominant model 0.99 (0.60–1.64), 1.11 (0.25–4.93), 0.85 (0.58–1.25),      
n=8 n=2 n = 4 (EA)
      CC vs AA 0.92 (0.40–2.09), 1.48 (0.78–2.80), 0.53 (0.15–1.90),      
n=7 n=2 n = 3 (EA)
Owen et al. MTHFR C677T Toxicity Dominant model 0.72 (0.47–1.11), 0.89 (0.55–1.45), 0.64 (0.34–1.20), No No [40]
(2013) n = 13 n=5 n = 8 (NC)
    Efficacy Dominant model 0.95 (0.76–1.11), 1.03 (0.73–1.47), 0.85 (0.57–1.23),      
n = 10 n=5 n = 5 (NC)
  MTHFR A1298C Efficacy Dominant model 1.23 (0.81–1.89), 1.01 (0.65–1.56), 1.59 (0.75–3.33), No No  
n=8 n=4 n = 4 (NC)
    Toxicity Dominant model 0.84 (0.56–1.25), 0.67 (0.38–1.19), 1.02 (0.59–1.75),      
n=8 n=3 n = 5 (NC)
Spyridopoulou MTHFR C677T Toxicity CT vs CC 1.57 (0.76–3.24), NA NA NA Yes [92]
et al. (2012) n=6
      TT vs CC 1.99 (0.98–4.04), NA NA      
n=6
EA: East Asian; NA: Not acquired; NC: Non-Caucasian; NOA: Non-Asian.

future science group

 Table 3. Associations between genetic polymorphisms and efficacy and/or toxicity of methotrexate in patients with rheumatoid arthritis reported by
previously published meta-analyses (cont.).
Study (year)  Genetic Outcomes Genetic model Overall patients Caucasians Asians Association with Ref.
polymorphism  (OR [95% CI], (OR [95% CI], (OR [95% CI], methotrexate
    study number)  study number) study number)  Efficacy Toxicity  

future science group

Spyridopoulou MTHFR A1298C Toxicity AC vs AA 1.49 (0.81–2.74), NA NA NA Yes  
et al. (2012; n=3
cont.)     CC vs AA 0.54 (0.19–1.57), NA NA      
  n=3
Lee et al. MTHFR C677T Efficacy Dominant model 1.18 (0.83–1.67), NA NA No No [93]
(2010) n=3
    Toxicity Dominant model 1.59 (0.81–3.03), 1.49 (0.83–2.70), 1.61 (0.60–4.35),      
n=6 n = 2 (NOA) n=4
  MTHFR A1298C Efficacy Dominant model 0.93 (0.43–2.04), NA NA No No  
n=2
    Toxicity Dominant model 1.06 (0.54–2.08), 1.04 (0.20–5.26), 1.02 (0.60–1.75),      
n=4 n = 2 (NOA) n=2
Fisher et al. MTHFR C677T Toxicity Dominant model 1.60 (0.90–2.86), NA NA Not Not [94]
(2009) n=8 given given
  MTHFR A1298C Toxicity Dominant model 0.83 (0.05–1.26), NA NA      
n=5
EA: East Asian; NA: Not acquired; NC: Non-Caucasian; NOA: Non-Asian.
Efficacy & toxicity of methotrexate 

www.futuremedicine.com
189
Systematic Review
Systematic Review  Chen, Zou, Sun, Yang & Liu
cacy and toxicity were unable to pool data [16,20,25,34–
37,44,49,67,69,72–74] . For efficacy, MTHFR C677T, TYMS
1494del6 and AMPD1 34C/T polymorphisms were
associated with responsiveness while ATIC T675C,
ATIC 347C/G and ITPA 94C/A polymorphisms were
related to nonresponsiveness [34,35,67,69,72] . Other gene
polymorphisms (e.g., MTHFR 1298 A/C) did not
have relationship with response (see Supplementary
Table 3 ).
For toxicity, polymorphisms in MTHFR C677T,
ATIC 347C/G and TYMS 2R/3R were related to
presenting overall adverse events while in GGH 401
C>T and SLCO1B1 T>C (rs4149056) were associ-
ated with absenting overall adverse events [20,37,44,72] .
Polymorphisms in MTHFR A1298C was associ-
ated with presenting gastrointestinal adverse events
while in SLCO1B1 T>C (rs4149056) and SLC19A1
G>A (rs7499) were associated with absenting gastro­
intestinal adverse events [37,73] . Genetic variations
in SLC19A1 G>A (rs1051266) was associated with
absenting gastro­intestinal adverse events, infections
and dermatologic adverse events and less frequencies of
discontinuation due to adverse events [37,49] . Polymor-
phisms in SLC19A1 were also associated with absenting
adverse events (SLC19A1 rs1131596 T/C for discontin-
uation due to adverse events; SLC19A1 rs2838956 T/C,
SLC19A1 rs1131596 T/C and SLC19A1 rs2838958 T/C
for infections; SLC19A1 rs2838956 T/C, SLC19A1
rs1131596 T/C and SLC19A1 rs2838958 T/C for
dermatologic adverse events) [49] . Poly­morphisms
in other genes (e.g., miR-499a rs3746444 A/G)
were not related to MTX-treated adverse events (see
Supplementary Table 3).
Sensitivity analysis & publication bias
We intended to perform sensitive analysis by pooling
the unadjusted ORs reported by the included studies
that didn’t report raw data to those calculated from raw
data. However, this could not be performed because
of the small study number that only two reported
­unadjusted ORs.
Funnel plots of association between MTHFR C677T
polymorphism and efficacy and overall adverse events
of MTX in RA patients in dominant genetic model
were symmetry, suggesting low risk of publication bias
(Supplementary Figures 1 & 2) .
Discussion
MTX enters into cells by the transportation of RFC1,
also called SLC19A1, and is pumped out through the
ABC, also called MDR. Within cells, MTX will be
translated to MTX polyglutamates (MTX-PGs), which
is reported to have association with therapy response
in RA patients [84] , by FPGS and MTX-PG will be
190 Pharmacogenomics (2017) 18(2)
subsequently cleaved by GGH. The more stable forms
of MTX, MTX-PGs, act as inhibitors that suppress
various enzymes, TYMS, dihydrofolate reductase and
ATIC, which involved in different metabolism path-
ways, for example, purine and pyrimidine metabolism
to inhibit DNA synthesis [13,85,86] .
Genetic polymorphisms in these enzymes and
transporters may affect treatment outcomes of MTX
in RA patients. Our study synthesized currently the
best evidence as regard to the associations between
gene polymorphisms and the efficacy/toxicity of MTX
treatment in patients with RA using a systematic and
comprehensive approach, finding AMPD1 34C>T,
ATIC T675C and SLC19A1/RFC-1 80G/A were asso-
ciated with treatment response while MTHFR C677T
was ­correlated to presenting adverse events.
Findings of previously published meta-analyses
[32,40,87–94] were summarized in Table 3 and most of
the evidence was focused on MTHFR C677T and
MTHFR A1298C polymorphisms which were assessed
in six studies. Polymorphisms in RFC1 80G/A were
related to efficacy while in MTHFR C677T were
associated with toxicity in MTX-treated RA patients,
which are consistent with our findings. Neverthe-
less, our study found additionally that TYMS 28 bp
(2R/3R) and TYMS 1494 del6 polymorphisms were
related to absenting overall adverse events in non-­
Caucasians and MTHFR A1298C polymorphisms
were not ­correlated with toxicity.
MTHFR C677T and MTHFR A1298C poly­
morphisms were the most studied genetic variants and
known to lead to decreased enzyme activity that might
be the potential cause of MTX-associated hepato­
toxicity in RA patients [95] . Our study found a corre-
lation between MTHFR C677T polymorphisms and
overall adverse events, but no associations with adverse
events of specific systems. Moreover, no relation-
ship between MTHFR A1298C polymorphisms and
adverse events was found.
There had some inconsistencies between our pooled
ORs using raw data and those ORs adjusted for poten-
tial confounders reported in single included studies.
For example, polymorphisms in, such as, MTHFR 677
C/T, TYMS 1494del6, ITPA 94C/A and ATIC 347C/G
were reported to be related to efficacy, and MTHFR
1298 A/C, GGH 401 C>T, SLCO1B1 T>C and
SLC19A1 were reported to be associated with adverse
events after adjusting for potential covariates such as
age and gender, which were not confirmed in our study.
However, comparisons between pooled adjusted ORs
and pooled unadjusted ORs were not applicable in this
systematic review for the lack of ­eligible data.
This study has some limitations. First, we were not
able to pool the ORs reported by the included studies
future science group
 Efficacy & toxicity of methotrexate  Systematic Review

that didn’t report raw data to those calculated from raw
data because of limited number of the former studies.
However, the results from un-pooled ORs suggested an
association between ATIC C347G and presenting over-
all and gastrointestinal adverse events, MTHFR C677T
with presenting central nervous system adverse events,
and SHMT1 C1420T and TSER*2/*3 with absent-
ing alopecia, which were reported by only one study,
respectively. Further studies are needed to confirm
these findings. Second, the quality of include studies
was not assessed for the lack of proper quality assess-
ment tool [96] , which may compromise the ­reliability
of the results.
Conclusion & future perspective
In conclusion, AMPD1 34C>T and ATIC T675C in
overall population and Caucasians, and SLC19A1/
RFC-1 80G/A in Caucasians are related to responsive-
ness while TYMS 28 bp (2R/3R) in non-­Caucasians
and ABCB1/MDR-1 C3435T in Caucasians are
associated with nonresponsiveness of MTX treat-
ment. MTHFR C677T in overall population and
non-­Caucasians is associated with presenting overall
adverse events while TYMS 1494 del6 in overall popu-
lation and non-Caucasians is correlated to absenting
adverse events. What is more, TYMS 28 bp (2R/3R)
relates to presenting adverse events in Caucasians but
absenting adverse events in non-Caucasians. Limited
evidence of adjusted ORs shows polymorphisms in
SLC19A1 is associated with absenting gastrointestinal
adverse events, infections, dermatologic adverse events
and lower rate of w
­ ithdraw due to adverse events.
To detect treatment predictors of MTX in RA to
conduct individualized therapy with maximum effi-
cacy but minimum adverse events is quite challeng-
ing, as MTX-related treatment has a multifactorial
ground, implicating a combination of patient related,
disease related, treatment related, and gene related
factors. In addition to genetic predictors, evidence of
epigenetic predictors is accumulating [97] , which may
be equally important to predict treatment outcomes
in MTX-treated RA patients in the future. Selecting
combination factors (e.g., gene–gene interactions [98])
of predictors is highly recommended in future studies
and using those predictors to carry out personalized
therapy and accumulate treatment experience are also
encouraged.
Supplementary data
To view the supplementary data that accompany this paper,
please visit the journal website at: www.futuremedicine.com/
doi/full/10.2217/pgs-2016-0158
Author contributions
G Liu is the joint senior author. G Liu and Y Chen conceived
the study. Y Chen designed the study forms and searched
the literature. Y Chen, K Zou, J Sun and Y Yang screened
the studies for inclusion, extracted data. Y Chen and
K  Zou ­analyzed data. All authors drafted and revised the
­manuscript.
Financial & competing interests disclosure
Y Chen was funded by China Scholarship Council (grant no.
201606240021). The authors have no other relevant affilia-
tions or financial involvement with any organization or entity
with a financial interest in or financial conflict with the subject
matter or materials discussed in the manuscript apart from
those disclosed.
No writing assistance was utilized in the production of this
manuscript.

Executive summary
Background
• Methotrexate (MTX) has been the most widely used anchor disease-modifying antirheumatic drug in patients
with rheumatoid arthritis (RA) with definitely efficacy as well as noticeable toxicity.
• Identifying the predictors of responsiveness and adverse events in MTX-treated RA patients has been the
focus of most concern, but still without consistent consensus.
Association between gene polymorphisms & efficacy in MTX-treated RA patients
• AMPD1 34C>T and ATIC T675C polymorphisms were correlated to treatment response in overall population.
• Polymorphisms in AMPD1 34C>T and in SLC19A1/RFC-1 80G/A was associated with response in Caucasians
while in TYMS 28 bp (2R/3R) related to nonresponse in non-Caucasians.
Association between gene polymorphisms & toxicity in MTX-treated RA patients
• Polymorphisms in TYMS 1494 del6 and in FPGS rs10106 were related to absenting overall adverse events in
overall patients.
• MTHFR C677T polymorphism had association with presenting overall adverse events in overall patients and in
non-Caucasians, but no statistical results found in MTHFR A1298C polymorphism.
• TYMS 28 bp (2R/3R) was correlated to present overall adverse events in Caucasians while associated with
absenting overall adverse events in non-Caucasians. ABCB1/MDR-1 C3435T related to absent overall adverse
events in Caucasians.

future science group www.futuremedicine.com 191

Systematic Review  Chen, Zou, Sun, Yang & Liu
References
1 Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis.
Lancet 388(10055), 2023–2038 (2016).
http://dx.doi.org/10.1016/S0140-6736(16)30173-8.
2 Carmona L, Cross M, Williams B, Lassere M, March L.
Rheumatoid arthritis. Best Pract. Res. Clin. Rheumatol. 24(6),
733–745 (2010).
3 Cross M, Smith E, Hoy D, Carmona L. The global burden
of rheumatoid arthritis: estimates from the global burden of
disease 2010 study. Ann. Rheum. Dis. 73(7), 1316–22 (2014). 
4 Usenbo A, Kramer V, Young T, Musekiwa A. Prevalence of
arthritis in Africa: a systematic review and meta-analysis.
PLoS ONE 10(8), e0133858 (2015).
5 Sokka T, Kautiainen H, Toloza S et al. QUEST-RA:
quantitative clinical assessment of patients with rheumatoid
arthritis seen in standard rheumatology care in 15 countries.
Ann. Rheum. Dis. 66(11), 1491–1496 (2007).
6 Pincus T, Gibson KA, Castrejon I. Update on methotrexate
as the anchor drug for rheumatoid arthritis. Bull. Hosp. Jt
Dis. 71(Suppl. 1), S9–S19 (2013).
7 Lopez-Olivo MA, Siddhanamatha HR, Shea B, Tugwell P,
Wells GA, Suarez-Almazor ME. Methotrexate for treating
rheumatoid arthritis. Cochrane Database Syst. Rev. 6,
CD000957 (2014).
8 Romao VC, Lima A, Bernardes M, Canhao H, Fonseca
JE. Three decades of low-dose methotrexate in rheumatoid
arthritis: can we predict toxicity? Immunol. Res. 60(2–3),
289–310 (2014).
9 Hornung N, Ellingsen T, Attermann J, Stengaard-Pedersen
K, Poulsen JH. Patients with rheumatoid arthritis treated
with methotrexate (MTX): concentrations of steady-state
erythrocyte MTX correlate to plasma concentrations and
clinical efficacy. J. Rheumatol. 35(9), 1709–1715 (2008).
10 Dalrymple JM, Stamp LK, O’Donnell JL, Chapman
PT, Zhang M, Barclay ML. Pharmacokinetics of oral
methotrexate in patients with rheumatoid arthritis. Arthritis
Rheum. 58(11), 3299–3308 (2008).
11 Bansard C, Lequerre T, Daveau M et al. Can rheumatoid
arthritis responsiveness to methotrexate and biologics be
predicted? Rheumatology (Oxford). 48(9), 1021–1028 (2009).
12 Kremer JM. Methotrexate pharmacogenomics. Ann. Rheum.
Dis. 65(9), 1121–1123 (2006).
13 Tarnowski M, Paradowska-Gorycka A, Dabrowska-Zamojcin
E, Czerewaty M, Sluczanowska-Glabowska S, Pawlik A.
The effect of gene polymorphisms on patient responses to
rheumatoid arthritis therapy. Expert Opin. Drug Metab.
Toxicol. 12(1), 41–55 (2016).
14 Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred
reporting items for systematic reviews and meta-analyses: the
PRISMA statement. Int. J. Surg. 8(5), 336–341 (2010).
15 Higgins JPT, Green Se. Cochrane Handbook for
Systematic Reviews of Interventions Version 5.1.0 [updated
March 2011]. Chapter 9: Heterogeneity. The Cochrane
Collaboration 2011; The Cochrane Collaboration (2011).
www.cochrane-handbook.org.
16 Toraih EA, Ismail NM, Toraih AA, Hussein MH, Fawzy
MS. Precursor miR-499a variant but not miR-196a2 is
192 Pharmacogenomics (2017) 18(2)
associated with rheumatoid arthritis susceptibility in an
Egyptian population. Mol. Diagn. Ther. 20(3), 279–295
(2016).
17 Muralidharan N, Mariaselvam CM, Jain VK, Gulati R, Negi
VS. ATIC 347C>G gene polymorphism may be associated
with methotrexate-induced adverse events in south Indian
Tamil rheumatoid arthritis. Pharmacogenomics 17(3),
241–248 (2016).
18 Muralidharan N, Mariaselvam CM, Cb M, Negi VS.
Reduced folate carrier-1 80G > A gene polymorphism
is not associated with methotrexate treatment response
in South Indian Tamils with rheumatoid arthritis. Clin.
Rheumatol. 35(4), 879–885 (2016).
19 Moya P, Salazar J, Arranz MJ et al. Methotrexate
pharmacokinetic genetic variants are associated with outcome
in rheumatoid arthritis patients. Pharmacogenomics 17(1),
25–29 (2016).
20 Swierkot J, Slezak R, Karpinski P et al. 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),
152–161 (2015).
21 Uribarri M, Ruiz-Larranaga O, Arteta D et al. Influence
of MTHFR C677T polymorphism on methotrexate
monotherapy discontinuation in rheumatoid arthritis
patients: results from the GAPAID European project. Clin.
Exp. Rheumatol. 33(5), 699–705 (2015).
22 Soukup T, Dosedel M, Pavek P et al. The impact of C677T
and A1298C MTHFR polymorphisms on methotrexate
therapeutic response in east Bohemian region rheumatoid
arthritis patients. Rheumatol. Int. 35(7), 1149–1161 (2015).
23 Saleh MM, Irshaid YM, Mustafa KN. Methylene
tetrahydrofolate reductase genotype frequencies: association
with toxicity and response to methotrexate in rheumatoid
arthritis patients. Int. J. Clin. Pharmacol. Ther. 53(2),
154–162 (2015).
24 Muralidharan N, Antony PT, Jain VK, Mariaselvam
CM, Negi VS. Multidrug resistance 1 (MDR1) 3435C>T
gene polymorphism influences the clinical phenotype
and methotrexate-induced adverse events in South Indian
Tamil rheumatoid arthritis. Eur. J. Clin. Pharmacol. 71(8),
959–965 (2015).
25 Lima A, Bernardes M, Azevedo R, Medeiros R, Seabra V.
Pharmacogenomics of methotrexate membrane transport
pathway: can clinical response to methotrexate in rheumatoid
arthritis be predicted? Int. J. Mol. Sci. 16(6), 13760–13780
(2015).
26 Kotrych D, Dziedziejko V, Safranow K, Pawlik A.
Lack of association between CXCL9 and CXCL10 gene
polymorphisms and the outcome of rheumatoid arthritis
treatment with methotrexate. Eur. Rev. Med. Pharmacol. Sci.
19(16), 3037–3040 (2015).
27 Iqbal MP, Ali AA, Mehboobali N, Iqbal K. Short
communication: lack of association between MTHFR gene
polymorphisms and response to methotrexate treatment in
Pakistani patients with rheumatoid arthritis. Pak. J. Pharm.
Sci. 28(5), 1789–1792 (2015).
future science group

28 Ghodke-Puranik Y, Puranik AS, Shintre P et al. Folate
metabolic pathway single nucleotide polymorphisms: a
predictive pharmacogenetic marker of methotrexate response
in Indian (Asian) patients with rheumatoid arthritis.
Pharmacogenomics 16(18), 2019–2034 (2015).
29 Chaabane S, Marzouk S, Akrout R et al. Genetic
determinants of methotrexate toxicity in tunisian patients
with rheumatoid arthritis: a study of polymorphisms
involved in the MTX metabolic pathway. Eur. J. Drug Metab.
Pharmacokinet. 41(4), 385–393 (2015).
30 Samara SA, Irshaid YM, Mustafa KN. Association of MDR1
C3435T and RFC1 G80A polymorphisms with methotrexate
toxicity and response in Jordanian rheumatoid arthritis
patients. Int. J. Clin. Pharmacol. Ther. 52(9), 746–755
(2014).
31 Prasad S, Tripathi D, Rai MK, Aggarwal S, Mittal B,
Agarwal V. Multidrug resistance protein-1 expression,
function and polymorphisms in patients with rheumatoid
arthritis not responding to methotrexate. Int. J. Rheum. Dis.
17(8), 878–886 (2014).
32 Morgan MD, Al-Shaarawy N, Martin S et al. MTHFR
functional genetic variation and methotrexate treatment
response in rheumatoid arthritis: a meta-analysis.
Pharmacogenomics 15(4), 467–475 (2014).
33 Mariaselvam CM, Sundaresh A, Chaaben AB et al.
Association of HLA-E*01:01/*01:03 polymorphism with
methotrexate-based treatment response in South Indian
rheumatoid arthritis patients. Indian J. Rheumatol. 9(4),
178–183 (2014).
34 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(10), e108165 (2014).
35 Lima A, Monteiro J, Bernardes M et al. Prediction of
methotrexate clinical response in Portuguese rheumatoid
arthritis patients: implication of MTHFR rs1801133 and
ATIC rs4673993 polymorphisms. Biomed. Res. Int. 2014,
368681 (2014).
36 Lima A, Bernardes M, Sousa H et al. SLC19A1 80G allele
as a biomarker of methotrexate-related gastrointestinal
toxicity in Portuguese rheumatoid arthritis patients.
Pharmacogenomics 15(6), 807–820 (2014).
37 Lima A, Bernardes M, Azevedo R et al. SLC19A1, SLC46A1
and SLCO1B1 polymorphisms as predictors of methotrexate-
related toxicity in Portuguese rheumatoid arthritis patients.
Toxicol. Sci. 142(1), 196–209 (2014).
38 Davis LA, Polk B, Mann A et al. Folic acid pathway single
nucleotide polymorphisms associated with methotrexate
significant adverse events in United States veterans with
rheumatoid arthritis. Clin. Exp. Rheumatol. 32(3), 324–332
(2014).
39 Pawlik A, Herczynska MM, Dziedziejko V et al. Effect of
allograft inflammatory factor-1 gene polymorphisms on
rheumatoid arthritis treatment with methotrexate. Postepy
Hig. Med. Dosw. (Online). 67, 637–642 (2013).
40 Owen SA, Lunt M, Bowes J et al. MTHFR gene
polymorphisms and outcome of methotrexate treatment
in patients with rheumatoid arthritis: analysis of key
future science group
Efficacy & toxicity of methotrexate  Systematic Review
polymorphisms and meta-analysis of C677T and A1298C
polymorphisms. Pharmacogenomics J. 13(2), 137–147 (2013).
41 Owen SA, Hider SL, Martin P, Bruce IN, Barton A,
Thomson W. Genetic polymorphisms in key methotrexate
pathway genes are associated with response to treatment in
rheumatoid arthritis patients. Pharmacogenomics J. 13(3),
227–234 (2013).
42 Jekic B, Lukovic L, Bunjevacki V et al. Association of the
TYMS 3G/3G genotype with poor response and GGH
354GG genotype with the bone marrow toxicity of the
methotrexate in RA patients. Eur. J. Clin. Pharmacol. 69(3),
377–383 (2013).
43 Stamp LK, Hazlett J, Roberts RL, Frampton C, Highton J,
Hessian PA. Adenosine receptor expression in rheumatoid
synovium: a basis for methotrexate action. Arthritis Res. Ther.
14(3), R138 (2012).
44 Plaza-Plaza JC, Aguilera M, Canadas-Garre M et al.
Pharmacogenetic polymorphisms contributing to toxicity
induced by methotrexate in the southern Spanish population
with rheumatoid arthritis. OMICS 16(11), 589–595 (2012).
45 Pawlik A, Dziedziejko V, Kurzawski M, Safranow K,
Kotrych D, Bohatyrewicz A. Effect of ESR1 and ESR2 gene
polymorphisms on rheumatoid arthritis treatment with
methotrexate. Pharmacol. Rep. 64(1), 185–190 (2012).
46 Milic V, Jekic B, Lukovic L et al. Association of dihydrofolate
reductase (DHFR) -317AA genotype with poor response to
methotrexate in patients with rheumatoid arthritis. Clin. Exp.
Rheumatol. 30(2), 178–183 (2012).
47 Choe JY, Lee H, Jung HY, Park SH, Bae SC, Kim SK.
Methylenetetrahydrofolate reductase polymorphisms, C677T
and A1298C, are associated with methotrexate-related toxicities
in Korean patients with rheumatoid arthritis. Rheumatol. Int.
32(6), 1837–1842 (2012).
48 Caliz R, del Amo J, Balsa A et al. 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), 10–14 (2012).
49 Bohanec Grabar P, Leandro-Garcia LJ, Inglada-Perez
L, Logar D, Rodriguez-Antona C, Dolzan V. Genetic
variation in the SLC19A1 gene and methotrexate toxicity
in rheumatoid arthritis patients. Pharmacogenomics 13(14),
1583–1594 (2012).
50 Tasbas O, Borman P, Gurhan Karabulut H, Tukun A,
Yorgancioglu R. The frequency of A1298C and C677T
polymorphisms of the methylentetrahydrofolate gene in
Turkish patients with rheumatoid arthritis: relationship with
methotrexate toxicity. Open Rheumatol. J. 5, 30–35 (2011).
51 Mena JP, Salazar-Paramo M, Gonzalez-Lopez L et al.
Polymorphisms C677T and A1298C in the MTHFR gene
in Mexican patients with rheumatoid arthritis treated with
methotrexate: implication with elevation of transaminases.
Pharmacogenomics J. 11(4), 287–291 (2011).
52 Xiao H, Xu J, Zhou X et al. Associations between the
genetic polymorphisms of MTHFR and outcomes of
methotrexate treatment in rheumatoid arthritis. Clin. Exp.
Rheumatol. 28(5), 728–733 (2010).
53 Stamp LK, Chapman PT, O’Donnell JL et al.
Polymorphisms within the folate pathway predict folate
www.futuremedicine.com 193
Systematic Review  Chen, Zou, Sun, Yang & Liu
concentrations but are not associated with disease
activity in rheumatoid arthritis patients on methotrexate.
Pharmacogenet. Genomics 20(6), 367–376 (2010).
54 Majorczyk E, Pawlik A, Kusnierczyk P. PTPN22 1858C>T
polymorphism is strongly associated with rheumatoid
arthritis but not with a response to methotrexate therapy. Int.
Immunopharmacol. 10(12), 1626–1629 (2010).
55 Kooloos WM, Wessels JA, van der Straaten T, Allaart CF,
Huizinga TW, Guchelaar HJ. Functional polymorphisms
and methotrexate treatment outcome in recent-onset
rheumatoid arthritis. Pharmacogenomics 11(2), 163–175
(2010).
56 Grabar PB, Rojko S, Logar D, Dolzan V. Genetic
determinants of methotrexate treatment in rheumatoid
arthritis patients: a study of polymorphisms in the adenosine
pathway. Ann. Rheum. Dis. 69(5), 931–932 (2010).
57 Taraborelli M, Andreoli L, Archetti S, Ferrari M, Cattaneo
R, Tincani A. Methylenetetrahydrofolate reductase
polymorphisms and methotrexate: no association with
response to therapy nor with drug-related adverse events
in an Italian population of rheumatic patients. Clin. Exp.
Rheumatol. 27(3), 499–502 (2009).
58 Stamp LK, O’Donnell JL, Chapman PT et al. Lack of
association between HLA-G 14 bp insertion/deletion
polymorphism and response to long-term therapy with
methotrexate response in rheumatoid arthritis. Ann. Rheum.
Dis. 68(1), 154–155 (2009).
59 Sharma S, Das M, Kumar A et al. Purine biosynthetic
pathway genes and methotrexate response in rheumatoid
arthritis patients among north Indians. Pharmacogenet.
Genomics 19(10), 823–828 (2009).
60 Lee YC, Cui J, Costenbader KH, Shadick NA, Weinblatt
ME, Karlson EW. Investigation of candidate polymorphisms
and disease activity in rheumatoid arthritis patients on
methotrexate. Rheumatology 48(6), 613–617 (2009).
61 Ranganathan P, Culverhouse R, Marsh S et al. Methotrexate
(MTX) pathway gene polymorphisms and their effects on
MTX toxicity in Caucasian and African American patients
with rheumatoid arthritis. J. Rheumatol. 35(4), 572–579
(2008).
62 James HM, Gillis D, Hissaria P et al. Common
polymorphisms in the folate pathway predict efficacy
of combination regimens containing methotrexate and
sulfasalazine in early rheumatoid arthritis. J. Rheumatol.
35(4), 562–571 (2008).
63 Hider SL, Thomson W, Mack LF, Armstrong DJ, Shadforth
M, Bruce IN. Polymorphisms within the adenosine
receptor 2a gene are associated with adverse events in RA
patients treated with MTX. Rheumatology (Oxford) 47(8),
1156–1159 (2008).
64 Ghodke Y, Chopra A, Joshi K, Patwardhan B. Are
Thymidylate synthase and Methylene tetrahydrofolate
reductase genes linked with methotrexate response
(efficacy, toxicity) in Indian (Asian) rheumatoid arthritis
patients? Clin. Rheumatol. 27(6), 787–789 (2008).
65 Bohanec Grabar P, Logar D, Lestan B, Dolzan V. Genetic
determinants of methotrexate toxicity in rheumatoid
arthritis patients: a study of polymorphisms affecting
194 Pharmacogenomics (2017) 18(2)
methotrexate transport and folate metabolism. Eur. J. Clin.
Pharmacol. 64(11), 1057–1068 (2008).
66 Wessels JA, van der Kooij SM, le Cessie S et al. A
clinical pharmacogenetic model to predict the efficacy of
methotrexate monotherapy in recent-onset rheumatoid
arthritis. Arthritis Rheum. 56(6), 1765–1775 (2007).
67 van der Straaten RJ, Wessels JA, de Vries-Bouwstra JK et al.
Exploratory analysis of four polymorphisms in human GGH
and FPGS genes and their effect in methotrexate-treated
rheumatoid arthritis patients. Pharmacogenomics 8(2),
141–150 (2007).
68 Taniguchi A, Urano W, Tanaka E et al. Validation of the
associations between single nucleotide polymorphisms or
haplotypes and responses to disease-modifying antirheumatic
drugs in patients with rheumatoid arthritis: a proposal for
prospective pharmacogenomic study in clinical practice.
Pharmacogenet. Genomics 17(6), 383–390 (2007).
69 Kurzawski M, Pawlik A, Safranow K, Herczynska M,
Drozdzik M. 677C>T and 1298A>C MTHFR polymorphisms
affect methotrexate treatment outcome in rheumatoid
arthritis. Pharmacogenomics 8(11), 1551–1559 (2007).
70 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(6), 404–407
(2007).
71 Berkun Y, Abou Atta I, Rubinow A et al. 2756GG genotype
of methionine synthase reductase gene is more prevalent
in rheumatoid arthritis patients treated with methotrexate
and is associated with methotrexate-induced nodulosis.
J. Rheumatol. 34(8), 1664–1669 (2007).
72 Wessels JA, Kooloos WM, De Jonge R et al. Relationship
between genetic variants in the adenosine pathway and
outcome of methotrexate treatment in patients with
recent-onset rheumatoid arthritis. Arthritis Rheum. 54(9),
2830–2839 (2006).
73 Wessels JA, de Vries-Bouwstra JK, Heijmans BT et al.
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), 1087–1095 (2006).
74 Takatori R, Takahashi KA, Tokunaga D et al. ABCB1
C3435T polymorphism influences methotrexate sensitivity
in rheumatoid arthritis patients. Clin. Exp. Rheumatol. 24(5),
546–554 (2006).
75 Rizzo R, Rubini M, Govoni M et al. HLA-G 14-bp
polymorphism regulates the methotrexate response in
rheumatoid arthritis. Pharmacogenet. Genomics 16(9),
615–623 (2006).
76 Kim SK, Jun JB, El-Sohemy A, Bae SC. Cost–effectiveness
analysis of MTHFR polymorphism screening by polymerase
chain reaction in Korean patients with rheumatoid arthritis
receiving methotrexate. J. Rheumatol. 33(7), 1266–1274
(2006).
77 Drozdzik M, Rudas T, Pawlik A et al. The effect of 3435C>T
MDR1 gene polymorphism on rheumatoid arthritis
treatment with disease-modifying antirheumatic drugs. Eur.
J. Clin. Pharmacol. 62(11), 933–937 (2006).
future science group

78 Aggarwal P, Naik S, Mishra KP, Aggarwal A, Misra R.
Correlation between methotrexate efficacy & toxicity with
C677T polymorphism of the methylenetetrahydrofolate gene
in rheumatoid arthritis patients on folate supplementation.
Indian J. Med. Res. 124(5), 521–526 (2006).
79 Berkun Y, Levartovsky D, Rubinow A et al. Methotrexate
related adverse effects in patients with rheumatoid arthritis
are associated with the A1298C polymorphism of the
MTHFR gene. Ann. Rheum. Dis. 63(10), 1227–1231 (2004).
80 Kumagai K, Hiyama K, Oyama T, Maeda H, Kohno
N. Polymorphisms in the thymidylate synthase and
methylenetetrahydrofolate reductase genes and sensitivity
to the low-dose methotrexate therapy in patients with
rheumatoid arthritis. Int. J. Mol. Med. 11(5), 593–600
(2003).
81 Urano W, Taniguchi A, Yamanaka H et al. Polymorphisms in
the methylenetetrahydrofolate reductase gene were associated
with both the efficacy and the toxicity of methotrexate used
for the treatment of rheumatoid arthritis, as evidenced by
single locus and haplotype analyses. Pharmacogenetics 12(3),
183–190 (2002).
82 van Ede AE, Laan RF, Blom HJ et al. The C677T mutation
in the methylenetetrahydrofolate reductase gene: a genetic
risk factor for methotrexate-related elevation of liver enzymes
in rheumatoid arthritis patients. Arthritis Rheum. 44(11),
2525–2530 (2001).
83 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. 54(2), 607–612 (2006).
84 den Boer E, de Rotte MC, Pluijm SM, Heil SG, Hazes JM,
de Jonge R. Determinants of erythrocyte methotrexate
polyglutamate levels in rheumatoid arthritis. J. Rheumatol.
41(11), 2167–2178 (2014).
85 Zhang LL, Yang S, Wei W, Zhang XJ. Genetic
polymorphisms affect efficacy and adverse drug reactions
of DMARDs in rheumatoid arthritis. Pharmacogenet.
Genomics 24(11), 531–538 (2014).
86 Abud-Mendoza C, Martinez-Martinez MU, Monsivais-
Urenda A, Gonzalez-Amaro R. Laboratory biomarkers for
guiding therapy with methotrexate in rheumatoid arthritis.
Curr. Pharm. Des. 21(2), 202–211 (2015).
87 Lee YH, Bae SC. Association of the ATIC 347 C/G
polymorphism with responsiveness to and toxicity of
methotrexate in rheumatoid arthritis: a meta-analysis.
Rheumatol. Int. 36(11), 1591–1599 (2016).
future science group
Efficacy & toxicity of methotrexate  Systematic Review
88 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. 38, 8–15
(2016).
89 Lee YH, Bae SC, Song GG. Association of the ABCB1
C3435T polymorphism with responsiveness to and toxicity
of DMARDs in rheumatoid arthritis: a meta-analysis.
Z Rheumatol. 75(7), 707–715 (2016).
90 Kung TN, Dennis J, Ma Y et al. RFC1 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(5), 1111–1120
(2014).
91 Song GG, Bae SC, Lee YH. Association of the MTHFR
C677T and A1298C polymorphisms with methotrexate
toxicity in rheumatoid arthritis: a meta-analysis. Clin.
Rheumatol. 33(12), 1715–1724 (2014).
92 Spyridopoulou KP, Dimou NL, Hamodrakas SJ, Bagos PG.
Methylene tetrahydrofolate reductase gene polymorphisms
and their association with methotrexate toxicity: a meta-
analysis. Pharmacogenet. Genomics 22(2), 117–133 (2012).
93 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. 30(2), 101–108 (2010).
94 Fisher MC, Cronstein BN. Metaanalysis of
methylenetetrahydrofolate reductase (MTHFR)
polymorphisms affecting methotrexate toxicity. J. Rheumatol.
36(3), 539–545 (2009).
95 te Loo DM, Hagleitner MM, Coenen MJ. Is there
a role for the MTHFR 677C>T and 1298A>C
polymorphisms in methotrexate-induced liver
toxicity? Pharmacogenomics 15(11), 1401–1403 (2014).
96 Zeng X, Zhang Y, Kwong JS et al. The methodological
quality assessment tools for preclinical and clinical studies,
systematic review and meta-analysis, and clinical practice
guideline: a systematic review. J. Evid. Based Med. 8(1), 2–10
(2015).
97 Plant D, Wilson AG, Barton A. Genetic and epigenetic
predictors of responsiveness to treatment in RA. Nat. Rev.
Rheumatol. 10(6), 329–337 (2014).
98 Dervieux T, Wessels JA, van der Straaten T et al. Gene–gene
interactions in folate and adenosine biosynthesis pathways
affect methotrexate efficacy and tolerability in rheumatoid
arthritis. Pharmacogenet. Genomics 19(12), 935–944 (2009).
www.futuremedicine.com 195