ORIGINAL ARTICLE

The role of clinical response to metformin in patients newlydiagnosed with type 2 diabetes: a monotherapy study

Abdolkarim Mahrooz • Hassan Parsanasab •

Mohammad Bagher Hashemi-Soteh • Zahra Kashi •

Adele Bahar • Ahad Alizadeh • Maliheh Mozayeni

Received: 31 January 2014 / Accepted: 26 March 2014

� Springer-Verlag Italia 2014

Abstract A major predicament in certain users of met-

formin, which is one of the most commonly used antihy-

perglycemic agents for type 2 diabetes (T2DM) treatment,

is the lack of appropriate response to the drug. We evalu-

ated the role of metformin response and OCT1 (organic

cation transporter1) Met420del polymorphism in a mono-

therapy study (metformin therapy for 12 weeks) on patients

newly diagnosed with T2DM. Based on the response to

metformin, patients (n = 108) were divided into two

groups: responders (n = 49) and non-responders (n = 59).

HbA1c levels were determined by affinity technique. The

OCT1-Met420del polymorphism was genotyped by PCR-

based restriction fragment length polymorphism. There

was a significant association between the variable response

with HbA1c and fasting blood sugar (FBS) (Wilks’

k = 0.905, p = 0.01). Responders had significantly lower

HbA1c and FBS levels compared with non-responders

(g2 = 0.087, p = 0.004 for HbA1c and g2 = 0.055,

p = 0.022 for FBS). The interaction treatment–response

increased the effect sizes from 32 to 58 % for HbA1c.

Alanine aminotransferase (ALT) and aspartate amino-

transferase (AST) values were significantly lower in the

responder group than in the non-responders (g2 = 0.067,

p = 0.01 for ALT and g2 = 0.052, p = 0.025 for AST).

This observational study showed that the variant OCT1-

Met420del may be more effective on plasma glucose than

HbA1c. The variable response could account for a signif-

icant proportion of the variance in HbA1c levels observed

following treatment with metformin. Metformin shows a

significantly greater effect on ALT and AST in responders

than in non-responders.

Keywords Metformin response � Type 2 diabetes � OCT �Met420del � Liver aminotransferase

Introduction

An epidemic increase in obesity, reduced physical activi-

ties, energy-rich diets, and overall negative changes in

lifestyle have led to the development of a number of dis-

orders such as type 2 diabetes mellitus (T2DM) and car-

diovascular diseases (CVD) [1, 2]. T2DM is a silent but

dangerous disease with a complex pathophysiology, and its

management has become a major health problem all over

the world [3, 4]. In Asia, compared with the Western

regions, the prevalence of T2DM has rapidly increased and

has affected the younger population [5, 6]. The prevalence

of this dangerous disease has been estimated to increase by

30–60 % by the year 2025 in several Asian countries [7].

Metformin is a biguanide derivative used widely as an

antihyperglycemic agent for the treatment of T2DM [8].

This magic drug is used as first-line therapy for T2DM, and

its hypoglycemic mechanisms include reduction of hepatic

A. Mahrooz (&) � H. Parsanasab � M. B. Hashemi-Soteh �M. Mozayeni

Department of Clinical Biochemistry and Genetics, Faculty

of Medicine, Mazandaran University of Medical Sciences,

Km 17 Khazarabad Road, Sari, Iran

e-mail: kmahrooz@yahoo.com; kmahrooz2@gmail.com

A. Mahrooz � M. B. Hashemi-Soteh

Molecular and Cell Biology Research Center,

Mazandaran University of Medical Sciences, Sari, Iran

Z. Kashi � A. Bahar

Diabetes Research Center, Imam Teaching Hospital,

Mazandaran University of Medical Sciences, Sari, Iran

A. Alizadeh

Department of Epidemiology and Biostatistics, Faculty

of Health, Tehran University of Medical Sciences, Tehran, Iran

123

Clin Exp Med

DOI 10.1007/s10238-014-0283-8

glucose production, enhanced glucose uptake and utiliza-

tion in skeletal tissue, and suppression of the intestinal

absorption of glucose [8, 9]. Investigations suggest that

metformin, in addition to T2DM, may prove beneficial in

the treatment of nonalcoholic fatty liver disease, polycystic

ovarian syndrome, CVD, and cancer [8, 10]. However, one

of the most important predicaments about metformin effi-

cacy is that a sufficient response is not observed in certain

patients receiving the drug [11]. Although drug disposition

and response could be influenced by various factors, such

as organ function and nature of the disease, estimates

indicate that genetics can dictate 20–95 % of variability in

different responses to the same medication [12].

Metformin, a hydrophilic organic cation, is a substrate

of at least two organic cation transporters (OCTs), namely

OCT1 and OCT2, which are responsible for hepatic dis-

tribution and renal excretion of metformin, respectively

[13, 14]. OCT1 as an uptake transporter plays a role in the

efficacy of metformin [15]. Polymorphisms in the OCT1

gene, particularly functional polymorphisms such as

Met420del, could affect metformin activity by influencing

its liver uptake, which in turn could influence the efficacy

of this important drug [13, 14]. The importance and role of

the OCTs and their genetic variations in metformin

metabolism suggest that further studies are needed to

examine the genetic variations and their relationships with

metformin response in different ethnic populations.

Moreover, there are contradictory reports on the associa-

tion between OCT1 variants and glucose-lowering

response to metformin [16].

Additionally, a number of studies have investigated the

effects of metformin therapy on ALT and AST in T2DM,

and the results are controversial [17–20]. One of the

parameters influencing these conflicting results may be that

metformin response has not been very well studied.

Therefore, one of the questions that the present study seeks

to answer is whether metformin response could regulate

changes in ALT and AST levels in patients newly diag-

nosed with T2DM.

Materials and methods

Patients

One hundred and eight patients with type 2 diabetes mel-

litus (age range 53.16 ± 9.7 years) treated with metformin

(1,000 mg per day) for 3 months were participated in the

study. All patients were newly diagnosed based on the

WHO criteria [21]. Based on the response to metformin,

patients were classified into two groups: responder group

(decrease in HbA1c levels by more than 1 % from the

baseline) and non-responder group (decrease in HbA1c

levels less than 1 % from the baseline). All patients

underwent a physical examination, and information about

demographic parameters, medical history, and medication

use was obtained using a questionnaire. None of the

patients were taking antidiabetic medication prior to their

diabetes diagnosis. Forty-eight patients were taking anti-

hypertensive medication such as losartan, an ACE inhibi-

tor, or a beta blocker, and most patients were receiving

lipid-lowering therapy. The study protocol was planned

based on the ethical criteria detailed in the Declaration of

Helsinki and was approved by the local ethics committee.

Informed consent was received from all participants.

Biochemical tests

The HbA1c levels were assayed by boronate affinity

technique (Axis-Shield PoC AS, Oslo, Norway; accuracy,

failure\5 %). Standard enzymatic tests were used to assay

values of fasting blood sugar (FBS), triglycerides (TGs),

total cholesterol (TC), HDL-C, ALT, and AST after an

overnight fast. The LDL-C values were calculated

according to the Friedewald method [22].

Identification of OCT1-Met420del variant

Genomic DNA was extracted from samples containing

EDTA by phenol–chloroform method [23]. The OCT1-

Met420del polymorphism was detected using the restricted

fragment length polymorphism (RFLP) analysis after PCR

amplification. The designed primers used for this poly-

morphism were F 50-AGGTTCACGGACTCTGTGCT-30

and R 50-AAGCTGGAGTGTGCGATCT-30. Genomic

DNA (400–500 ng) was amplified in 25 ll of reaction

mixture consisted of 220 lM of each dNTP, 280 nM of

each primer, 1.5 mM MgCl2 and 1 U Taq polymerase. PCR

was subjected to 93 �C for 3 min, followed by 35 cycles of

93 �C for 45 s, 58 �C for 35 s, and 72 �C for 35 s, with a

final extension step of 72 �C for 5 min. Amplification

products from each sample (600 bp) were cleaved by

BspHI (Fermentase, Lithuania) for 20 h and resulted in

197- and 403-bp fragments, which were subjected to

electrophoresis on a 1 % agarose gel. Wild-type (AA)

patients were identified by the presence of 600-bp frag-

ment. Heterozygous (Aa) patients were identified by the

presence of 600-, 197- and 403-bp fragments, while the

presence of 197- and 403-bp fragments was the basis for

identification of mutants (aa).

Statistical tests

Statistical calculations were performed by the softwares R

(version 3.0.1) and SPSS (version 16.0). The chi-square

test was used for testing Hardy–Weinberg equilibrium. The

Clin Exp Med

123

gene counting method was used to determine allele and

genotype frequencies. The differences of the nonparametric

variables were analyzed by Mann–Whitney test. Repeated

measures multivariate analysis of variance (repeated mea-

sures MANOVA) was used for testing the planned com-

parisons. Additional analyses were performed using the

repeated measures analysis of variance (repeated measures

ANOVA). The outcomes of equality test of the error

variances and equality test of the covariance matrices

showed that the analyses in the present study met the

assumptions underlying the repeated measures MANOVA

model. To indicate effect sizes, the Wilks’ lambda was

performed for the overall multivariate model and when the

model was significant, the Partial ETA squared (g2)

was performed for protected univariate F tests. A

p value \ 0.05 was accepted as statistically significant.

Results

In this monotherapy study, subjects fall into two groups:

responders (n = 49) and non-responders (n = 59). The

groups did not differ significantly in age (53.45 ± 9.48 in

the responder group, 52.95 ± 9.95 in the non-responder

group, p = 0.8). Of the 108 participants, 87 were women

(42 were responders and 45 were non-responders) and 21

were men (7 were responders and 14 were non-responders).

Values of the study parameters in baseline and after met-

formin therapy based on responders and non-responders are

presented in Table 1. As shown in the table, there was

statistically significant difference between responders and

non-responders after metformin therapy with respect to

FBS, HbA1c, ALT, and AST.

The allele frequency and genotype distribution of

OCT1-Met420del polymorphism are shown in Table 2.

The genotypes frequencies were 66.7 % AA, 29.2 % Aa,

4.2 % aa in responder group and 49.2 % AA, 44.1 % Aa,

6.8 % aa in non-responder group. The Aa genotype and aa

were overrepresented among non-responder group in

comparison with responder group, resulting in a higher

frequency of the mutant allele Met420del in non-respond-

ers than in responders (0.29 vs. 0.19; p = 0.088). In

responders, the prevalence of AA genotype (OR = 2.07,

p = 0.069) was found to be higher than non-responders.

Our statistical models denoted the following as inde-

pendent variables: response, Met420del variant, sex,

treatment (metformin therapy), and their interactions such

as treatment–response, treatment–sex and Met420del–

response, while HbA1c, FBS, ALT, and AST were con-

sidered as dependent variables. The multivariate analysis

(Table 3) identified a significant association between the

variables response and treatment with HbA1c and FBS

(Wilks’ k = 0.905, F = 4.82, p = 0.01 for response and

Wilks’ k = 0.627, F = 27.39, p \ 0.001 for treatment).

Our findings showed that after metformin therapy, HbA1c

and FBS levels were significantly lower compared with

before treatment (g2 = 0.318, F = 43.39, p \ 0.001 for

HbA1c and g2 = 0.216, F = 25.61, p \ 0.001 for FBS). In

addition, the responders had significantly lower HbA1c and

Table 1 Levels of the study

parameters in baseline and after

metformin therapy according to

responders and non-responders

(n = 108)

Responder group: decrease in

HbA1c levels by more than 1 %

from the baseline, and non-

responder group: decrease in

HbA1c levels less than 1 % from

the baseline

Data were analyzed by Mann–

Whitney test

Parameter Baseline After 12 weeks

Non-responders Responders p value Non-responders Responders p value

SBP (mmHg) 133.85 ± 15.13 128.75 ± 15.93 0.095 127.73 ± 13.22 120.48 ± 22.14 0.018

DBP (mmHg) 82.37 ± 10.19 78.75 ± 9.97 0.039 78.59 ± 10.58 76.25 ± 9.76 0.082

FBS (mg/dL) 141.05 ± 22.28 148.19 ± 28.63 0.255 142.43 ± 42.78 116.66 ± 18.84 \0.001

HBA1C (%) 7.56 ± 0.69 7.97 ± 0.78 0.009 7.68 ± 1.15 6.30 ± 0.78 \0.001

ALT (U/L) 25.14 ± 8.58 23.54 ± 10.78 0.094 26.85 ± 13.05 21.65 ± 7.06 0.026

AST (U/L) 25.18 ± 11.62 22.33 ± 10.6 0.178 26.88 ± 13.39 21.04 ± 8.66 0.012

TG (mg/dL) 172.95 ± 82.11 180.69 ± 65.02 0.256 160.32 ± 61.78 154.38 ± 57.27 0.612

TC (mg/dL) 183.34 ± 45.87 186.96 ± 36.57 0.476 176.19 ± 36.23 171.08 ± 29.01 0.766

HDL-C (mg/dL) 48.29 ± 15.97 45.46 ± 16.28 0.237 48.24 ± 14.23 51.79 ± 17.46 0.293

LDL-C (mg/dL) 97.54 ± 32.93 106.54 ± 37.45 0.208 89.05 ± 23.33 87.71 ± 27.26 0.975

Table 2 Allele frequencies and genotype distribution of OCT1-

Met420del polymorphism in non-responders and responders

Non-responder

(n = 59)

Responder

(n = 48)

OR 95 % CI p

Genotype, n (%)

AA 29 (49.2) 32 (66.7) 2.069a 0.941–4.549 0.069

Aa 26 (44.1) 14 (29.2) 0.523b 0.233–1.171 0.113

aa 4 (6.8) 2 (4.2) 0.598c 0.105–3.413 0.639

Allele frequency

A 84 (71.2) 78 (81.3)

a 34 (28.8) 18 (18.8) 0.570d 0.298–1.09 0.088

a AA versus (Aa ? aa)b Aa versus (AA ? aa)c aa versus (AA ? Aa)d a versus A

Clin Exp Med

123

FBS levels than those of non-responders (g2 = 0.087,

F = 8.82, p = 0.004 for HbA1c and g2 = 0.055,

F = 5.43, p = 0.022 for FBS). The Partial ETA squared

(g2) represents the proportion of the variance in the

dependent variables that can be explained using the inde-

pendent variables. With respect to treatment as an inde-

pendent variable, the values obtained for g2 were 0.318 and

0.216 for HbA1c and FBS, respectively. In other words,

32 % of the variance in HbA1c and 22 % of the variance in

FBS can be explained by metformin therapy. When the

results of the interaction of treatment and response were

considered, the effect sizes increased up to 58 % for

HbA1c and up to 28 % for FBS.

When we analyzed the interaction effects between the 4

variables (response, treatment, Met420del and sex), a sig-

nificant interaction effect between treatment and response

was observed (Table 3; Wilks’ k = 0.401, F = 68.61,

p \ 0.001), as well as between treatment and sex (Table 3;

Wilks’ k = 0.898, F = 5.24, p = 0.007). In addition, a

borderline significant difference for the interaction effect

of treatment–Met420del was seen (Table 3; Wilks’

k = 0.943, F = 2.79, p = 0.067). This marginal differ-

ence in significance is primarily caused by the variation in

FBS levels among the diverse genotypes (Table 3;

g2 = 0.033, F = 3.14, p = 0.08). In other words, follow-

ing metformin therapy, the reduction of FBS concentra-

tions in wild-type individuals for the variant OCT1-

Met420del is greater than in the mutant allele carriers.

Figure 1 shows that the significant difference observed for

the interaction effect of treatment–sex is primarily due to

the changes in FBS levels (Table 3; g2 = 0.101,

F = 10.48, p = 0.002). Significant differences were not

observed for the other variables including sex, Met420del,

and Met420del–response (Table 3).

As indicated in Table 4, a significant relationship was

observed between the independent variables response with

ALT and AST (Wilks’ k = 0.932, F = 3.43, p = 0.037).

Separate analyses of each of the dependent variables dem-

onstrated statistically significant differences. The results

indicated that ALT and AST values were significantly lower

in the responder group than in the non-responders

(g2 = 0.067, F = 6.82, p = 0.01 for ALT and g2 = 0.052,

Table 3 The overall multivariate models and the related univariate

tests on the study dependent variables (HbA1c and FBS)

Independent

variable

Multivariate test Univariate tests

Dependent variables

HbA1c FBSb

Wilks’ k p value g2 p value g2 p value

Responsea 0.905 0.010 0.087 0.004 0.055 0.022

Treatment 0.627 0.000 0.318 0.000 0.216 0.000

Met420del 0.997 0.853 0.003 0.584 0.002 0.686

Sex 0.985 0.504 0.001 0.766 0.014 0.259

Treatment–sex 0.898 0.007 0.002 0.614 0.101 0.002

Treatment–

response

0.401 0.000 0.581 0.000 0.276 0.000

Treatment–

Met420del

0.943 0.067 0.012 0.291 0.033 0.080

a Response was defined as decrease in HbA1c levels by more than 1 %

from the baseline after 12 weeks of metformin therapyb FBS was transformed by natural logarithm

Fig. 1 The interaction effect of treatment–sex on fasting glucose

concentrations

Table 4 The overall multivariate models and the related univariate

tests on the study dependent variables (ALT and AST)

Independent

variable

Multivariate test Univariate tests

Dependent variables

ALT AST

Wilks’ k p value g2 p value g2 p value

Responsea 0.932 0.037 0.067 0.01 0.052 0.025

Treatment 0.999 0.971 0.000 0.974 0.000 0.851

Met420del 0.911 0.012 0.034 0.07 0.000 0.98

Sex 1.000 0.985 0.000 0.864 0.000 0.902

Treatment–sex 0.989 0.592 0.002 0.644 0.01 0.336

Treatment–

response

0.99 0.627 0.008 0.379 0.009 0.358

Treatment–

Met420del

0.993 0.706 0.006 0.45 0.007 0.45

a Response was defined as decrease in HbA1c levels by more than 1 %

from the baseline after 12 weeks of metformin therapy

Clin Exp Med

123

F = 5.16, p = 0.025 for AST). Although there was a sig-

nificant link between Met420del and the aminotransferases

(Wilks’ k = 0.911, F = 4.61, p = 0.012), statistically sig-

nificant differences were not found (g2 = 0.034, F = 3.36,

p = 0.07 for ALT and g2 \ 0.001, F = 0.001, p = 0.98 for

AST). There were no significant differences between the

other variables: sex, treatment, treatment–Met420del,

treatment–sex, and treatment–response (Table 4). It should

be noted that the interactions that were not statistically sig-

nificant, such as treatment–response–Met420del, are not

reported.

Discussion

Owing to the variability in response to metformin,

approximately 35–40 % of patients taking this important

drug do not achieve the expected glucose-lowering

response [13, 24]. It has been demonstrated that the OCTs

play a dominant role in glycemic response or renal elimi-

nation of metformin response [13]. Polymorphisms in the

OCT1 gene, particularly the variants with reduced func-

tions, contribute to the variations in response to metformin

[14]. Because of the multifactorial nature of metformin

response and the controversial reports on pharmacological

mechanisms of metformin [25, 26], additional investiga-

tions are required to identify the ethnic variability in the

OCT1 gene and the interindividual differences in response

to the drug.

Some studies have reported that there are no accepted

criterion for grouping the diabetic patients into responders

and non-responders [27]. Shikata et al. [27] selected

reduction of HbA1c values by more than 0.5 % as a cutoff

point for dividing patients into responders and non-

responders. The selection of the researchers was based on

their clinical experiences, but it was reported, in a sys-

tematic review, that after 3 months of metformin therapy

(doses up to 1,500 mg/day) HbA1c values decreased by

approximately 1 % compared with placebo [28]. Thus, in

the present study, a reduction of C1 % in HbA1c was

deemed a response to metformin therapy.

In our study, the frequency of the mutant allele of

OCT1-Met420del variant was higher in non-responders

than in responders, and this might influence the lower FBS

and HbA1c levels observed in non-responders. According

to our findings, which concurred with previous reports,

subjects with the mutant allele (risk allele) presented

increased glucose levels compared with those without the

allele, although the results were not statistically significant.

The mutant allele carriers might have a reduced metformin

uptake into the liver, resulting in decreased effects of

metformin in the organ. Based on this hypothesis, a con-

sequence of reduced hepatic metformin uptake is a

decreased hepatic glucose production, which is the most

important mechanism of the drug action [9], and this in

turn may affect plasma glucose. These observations are

partially in accordance with the studies by Becker et al.

[29], wherein metformin was found less effective in low-

ering HbA1c and glucose levels in individuals expressing

the mutant allele because of a functional polymorphism in

the OCT1 gene than those expressing the wild-type gene.

According to the study by Zhou et al. [26], the variant

Met420del did not influence the initial HbA1c reduction

following metformin therapy, which agreed with our

observations that carriers of the OCT1-Met420del poly-

morphism consistently did not show a significant change in

the ability of metformin to lower HbA1c. The low-activity

variations of OCT1 seem more effective in reducing

plasma glucose than that of HbA1c, although studies with

larger sample sizes will be necessary to provide more

definite conclusions. In other words, effect of the

Met420del may not be reflected in the long-term glycemic

control, but it may be effective in regulating plasma glu-

cose concentrations. The limitation of our present study

was that we did not assay plasma metformin concentrations

to examine the link between the metformin levels and the

allelic and genotypic distribution of the Met420del variant.

However, our observations may be confirmed by the find-

ing that the risk allele carriers of some OCT1 polymor-

phisms, including those carrying Met420del, had a higher

plasma metformin concentration than subjects with the

wild-type allele [13]. In general, considering the different

consequences of the variations in drug transporters on their

gene expression, mRNA stability, degradation, protein

folding, and substrate binding [30], the effect of OCT1

variations on metformin cannot be ignored.

Previous studies have shown that oral antidiabetic drugs

effectively decrease HbA1c levels by 0.5–1.5 % [28]. Our

study indicated that following metformin therapy, the

average decrease in HbA1c levels reached 0.68 %

(0.68 % ± 1.18) in all participants. In addition, the mean

decrease was 0.67 % (0.67 ± 0.58 higher than the criterion

HbA1c level) among responders, whereas the value was

-0.12 % (-0.12 ± 0.9) in non-responders. This result

demonstrates that the greater proportion of the decrease in

HbA1c levels in all patients after metformin therapy was

associated with responders, suggesting that metformin

response could be important in evaluating HbA1c as a key

indicator in monitoring the long-term glycemic control.

Shikata et al. [27] consistently showed that responders to

metformin had significantly lower HbA1c values compared

with non-responders. The results were further supported by

the findings that the interaction effects between the vari-

ables treatment and response increased the effective size

from 32 to 58 % for HbA1c, which is the largest effect

observed on HbA1c. A g2 value of 0.581 suggests that

Clin Exp Med

123

58 % of the variation in HbA1c values was caused by the

interaction between treatment–response. On the other hand,

the variable response accounts for 26 % of the greater

variance in HbA1c levels.

In the present study, the effect of metformin on HbA1c

and FBS, in terms of sex, was considered. The interaction

treatment–sex demonstrated that these two factors do not

have a simultaneous effect on HbA1c, but metformin

therapy leads to a greater reduction of glucose levels in

males than in females. In other words, metformin may be

more effective in men compared with women. The results

reported by Orchard et al. [31] partially agreed with our

study and indicated that metformin was significantly

effective in men compared with placebo, but was not

effective in women. Moreover, Zhou et al. [26] reported

that sex has no clinical effect on the response to metformin.

Our study also showed that the interaction sex–response

had no effect on HbA1c and glucose levels.

Drug treatment studies demonstrated that glitazones

decrease liver aminotransferases levels particularly ALT,

whereas reports on metformin are variable [17, 20, 32].

While some studies have reported a significant reduction in

ALT and AST levels after metformin therapy in nonalco-

holic steatohepatitis [17, 18], others have indicated that the

levels of the enzyme remain unchanged after metformin

treatment in T2DM [19, 20]. Our findings may be useful in

explaining these contradictory reports. According to our

results, there were no significant differences between

before and after metformin treatment, with respect to ALT

and AST, which is in accordance with other authors [19,

20]. Additional analyses with response indicated that

metformin is significantly more effective on ALT and AST

levels in responders than non-responders. Changes in the

liver aminotransferase levels by metformin could be con-

sidered as potential markers for evaluating the hepatic

function in metformin therapy [3, 20]. Therefore, it is of

importance to consider metformin response as an effective

factor in evaluating the drug effect on ALT and AST values

particularly in T2DM. Previous studies suggest that

decreased liver aminotransferase levels after metformin

therapy may be a reflection of decreased hepatic fat

accumulation, although the precise mechanism is unknown

[32, 33]. Our results generally provide the basis for basic

studies to elucidate mechanisms by which response to

metformin affects liver aminotransferases.

This is an observational study; thus, we are unable to

control certain situations. Some of those who participated

in the study were taking antihypertensive medications and

statin. In patients who were taking antihypertensive drugs

or statin, the responders and non-responders were signifi-

cantly different in FBS and HbA1c levels. Although these

drugs on FBS and HbA1c levels may be effective, in those

who did not take these drugs, there was still a significant

difference between responders and non-responders. Indeed,

there was almost no difference between responders and

non-responders in the number of people who were taking

antihypertensive drugs or statin therapy: Of the 48 patients

who received antihypertensive drugs, 24 were responders

and 24 were non-responders; of the 76 patients who

received statin, 33 were responders and 43 were non-

responders.

In this study, two limitations should be considered: the

relatively small sample size and the short duration of the

study.

Conclusion

This study indicated that the OCT1-Met420del may be

more effective on plasma glucose than HbA1c. Decrease in

HbA1c levels by more than 1 % from baseline could be

considered a criterion for classifying diabetic patients as

responders or non-responders to metformin. The response

could account for a significant proportion of the variance in

HbA1c values following metformin therapy. Metformin

shows a significantly greater effect on ALT and AST in

responders than in non-responders, which may account for

the conflicting reports on metformin effects on the ami-

notransferases. Further investigations are warranted to

elucidate the related molecular mechanisms, particularly

interactions between loss-of-function variants in the genes

of metformin transporters.

Acknowledgments This research was funded with the support of

Mazandaran University of Medical Sciences, Sari, Iran. The authors

thank Dr. Ozra Akha and Dr. Saeed Abedian Kenari for contributions.

Conflict of interest The authors have not declared any conflict of

interest.

References

1. Rizos CV, Elisaf MS (2013) Metformin and cancer. Eur J Phar-

macol 705:96–108

2. Salpeter SR, Buckley NS, Kahn JA, Salpeter EE (2008) Meta

analysis: metformin treatment in persons at risk for diabetes

mellitus. Am J Med 121:149–157

3. Zinman B, Harris SB, Neuman J, Gerstein HC, Retnakaran RR,

Raboud J et al (2010) Low-dose combination therapy with ros-

iglitazone and metformin to prevent type 2 diabetes mellitus

(CANOE trial): a double-blind randomised controlled study.

Lancet 376:103–111

4. Esteghamati A, Eskandari D, Mirmiranpour H, Noshad S,

Mousavizadeh M, Hedayati M et al (2013) Effects of metformin

on markers of oxidative stress and antioxidant reserve in patients

with newly diagnosed type 2 diabetes: a randomized clinical trial.

Clin Nutr 32:179–185

5. Gao Y, Yoon KH, Chuang L-M, Mohan V, Ning G, Shah S et al

(2009) Efficacy and safety of exenatide in patients of Asian

descent with type 2 diabetes inadequately controlled with

Clin Exp Med

123

metformin or metformin and a sulphonylurea. Diabetes Res Clin

Pract 83:69–76

6. Burcelin R, Serino M, Chabo C, Blasco-Baque V, Amar J (2011)

Gut microbiota and diabetes: from pathogenesis to therapeutic

perspective. Acta Diabetol 48:257–273

7. Chan JC, Deerochanawong C, Shera AS, Yoon K-H, Adam JM,

Binh TV et al (2007) Role of metformin in the initiation of

pharmacotherapy for type 2 diabetes: an Asian-Pacific perspec-

tive. Diabetes Res Clin Pract 75:255–266

8. Aljada A, Mousa SA (2012) Metformin and neoplasia: implica-

tions and indications. Pharmacol Ther 133:108–115

9. Ota S, Horigome K, Ishii T, Nakai M, Hayashi K, Kawamura T

et al (2009) Metformin suppresses glucose-6-phosphatase

expression by a complex I inhibition and AMPK activation-

independent mechanism. Biochem Biophys Res Commun

388:311–316

10. Viollet B, Guigas B, Garcia NS, Leclerc J, Foretz M, Andreelli F

(2012) Cellular and molecular mechanisms of metformin: an

overview. Clin Sci 122:253–270

11. Takane H, Shikata E, Otsubo K, Higuchi S, Ieiri I (2008) Poly-

morphism in human organic cation transporters and metformin

action. Pharmacogenomics 9:415–422

12. Kerb R (2006) Implications of genetic polymorphisms in drug

transporters for pharmacotherapy. Cancer Lett 234:4–33

13. DiStefano JK, Watanabe RM (2010) Pharmacogenetics of anti-

diabetes drugs. Pharmaceuticals 3:2610–2646

14. Holstein A, Seeringer A, Kovacs P (2011) Therapy with oral

antidiabetic drugs: applied pharmacogenetics. Br J Diabetes Vasc

Dis 11:10–16

15. Shu Y, Sheardown SA, Brown C, Owen RP, Zhang S, Castro RA

et al (2007) Effect of genetic variation in the organic cation

transporter 1 (OCT1) on metformin action. J Clin Invest

117:1422–1431

16. Nies AT, Koepsell H, Winter S, Burk O, Klein K, Kerb R et al

(2009) Expression of organic cation transporters OCT1

(SLC22A1) and OCT3 (SLC22A3) is affected by genetic factors

and cholestasis in human liver. Hepatology 50:1227–1240

17. Marchesini G, Brizi M, Bianchi G, Tomassetti S, Zoli M, Mel-

chionda N (2001) Metformin in non-alcoholic steatohepatitis.

Lancet 358:893–894

18. Schwimmer JB, Middleton MS, Deutsch R, Lavine JE (2005) A

phase 2 clinical trial of metformin as a treatment for non-diabetic

paediatric non-alcoholic steatohepatitis. Aliment Pharmacol Ther

21:871–879

19. Tiikkainen M, Hakkinen A-M, Korsheninnikova E, Nyman T,

Makimattila S, Yki-Jarvinen H (2004) Effects of rosiglitazone

and metformin on liver fat content, hepatic insulin resistance,

insulin clearance, and gene expression in adipose tissue in

patients with type 2 diabetes. Diabetes 53:2169–2176

20. Belcher G, Schernthaner G (2005) Changes in liver tests during

1-year treatment of patients with Type 2 diabetes with pioglit-

azone, metformin or gliclazide. Diabet Med 22:973–979

21. Bennet PH (1999) Impact of the new WHO classification and

diagnostic criteria. Diabetes Obes Metab 1(Suppl 2):S1–S6

22. Friedwald WT, Levy RI, Fredrickson DS (1972) Estimation of

the concentration of low-density lipoprotein cholesterol in

plasma, without use of the preparative ultracentrifuge. Clin Chem

18:499–502

23. Hoff-Olsena P, Mevag B, Staalstrøma E, Hovdea B, Egeland T,

Olaisen B (1999) Extraction of DNA from decomposed human

tissue: an evaluation of five extraction methods for short tandem

repeat typing. Forensic Sci Int 105:171–183

24. Pacanowski MA, Hopley CW, Aquilante CL (2008) Interindi-

vidual variability in oral antidiabetic drug disposition and

response: the role of drug transporter polymorphisms. Expert

Opin Drug Metab Toxicol 4:529–544

25. Zolk O (2012) Disposition of metformin: variability due to

polymorphisms of organic cation transporters. Ann Med

44:119–129

26. Zhou K, Donnelly LA, Kimber CH, Donnan PT, Doney AS,

Leese G et al (2009) Reduced-function SLC22A1 polymorphisms

encoding organic cation transporter 1 and glycemic response to

metformin: A GoDARTS study. Diabetes 58:1434–1439

27. Shikata E, Yamamoto R, Takane H, Shigemasa C, Ikeda T, Ot-

subo K et al (2007) Human organic cation transporter (OCT1 and

OCT2) gene polymorphisms and therapeutic effects of metfor-

min. J Hum Genet 52:117–122

28. Sherifali D, Nerenberg K, Pullenayegum E, Cheng JE, Gerstein

HC (2010) The effect of oral antidiabetic agents on A1C levels.

Diabetes Care 33:1859–1864

29. Becker ML, Visser LE, van Schaik RH, Hofman A, Uitterlinden

AG, Stricker BH (2009) Genetic variation in the organic cation

transporter 1 is associated with metformin response in patients

with diabetes mellitus. Pharmacogenomics J 9:242–247

30. Sissung TM, Troutman SM, Campbell TJ, Pressler HM, Sung H,

Bates SE et al (2012) Transporter pharmacogenetics: transporter

polymorphisms affect normal physiology, diseases, and pharma-

cotherapy. Discov Med 13:19–34

31. Orchard T, Temprosa M, Goldberg R, Haffner S, Ratner R,

Marcovina S et al (2005) The effect of metformin and intensive

lifestyle intervention on the metabolic syndrome: the diabetes

prevention program randomized trial. Ann Intern Med

142:611–619

32. Sattar N, McConnachie A, Ford I, Gaw A, Cleland SJ, Forouhi

NG et al (2007) Serial metabolic measurements and conversion to

type 2 diabetes in the west of Scotland coronary prevention study.

Diabetes 56:984–991

33. Matthews DR, Charbonnel BH, Hanefeld M, Brunetti P,

Schernthaner G (2005) Long-term therapy with addition of

pioglitazone to metformin compared with the addition of glicla-

zide to metformin in patients with type 2 diabetes: a randomized,

comparative study. Diabetes Metab Res Rev 21:167–174

Clin Exp Med

123