Comparison of vildagliptin-metformin and glimepiride-metformin treatments in type 2 diabetic patients.

Abstract

The present study investigated the efficacy and safety of vildagliptin-metformin treatment compared to those of glimepiride-metformin treatment for type 2 diabetes.
In a randomized, open-label, comparative study, 106 patients with type 2 diabetes were enrolled. The primary endpoint was a reduction in HbA1c from baseline and secondary endpoints included fasting plasma glucose (FPG) or 2-hour postprandial glucose (2h-PPG) reduction from baseline, as well as HbA1c responder rate and HbA1c reduction according to baseline HbA1c category.
Comparable HbA1c reduction was observed with a mean±standard deviation change from baseline to the 32-week endpoint of -0.94±1.15% in the vildagliptin group and -1.00±1.32% in the glimepiride group. A similar reduction in 2h-PPG (vildagliptin group 3.53±4.11 mmol/L vs. the glimepiride group 3.72±4.17 mmol/L) was demonstrated, and the decrements in FPG (vildagliptin group 1.54±2.41 mmol/L vs. glimepiride group 2.16±2.51 mmol/L) were not different between groups. The proportion of patients who achieved an HbA1c less than 7% at week 32 was 50.1% in the vildagliptin group and 56.0% in the glimepiride group. An average body weight gain of 2.53±1.21 kg in the glimepiride group was observed in contrast with the 0.23±0.69 kg weight gain noted in the vildagliptin group. A 10-fold lower incidence of hypoglycemia was demonstrated in the vildagliptin group, in addition to an absence of severe hypoglycemia.
Vildagliptin-metformin treatment provided blood glucose control efficacy comparable to that of glimepiride-metformin treatment and resulted in better adverse event profiles with lower risks of hypoglycemia and weight gain.

Full text

Page 1

D I A B E T E S & M E T A B O L I S M J O U R N A L
This is an Open Access article distributed under the terms of the Creative Commons At-
tribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/)
which permits unrestricted non-commercial use, distribution, and reproduction in any
medium, provided the original work is properly cited.
Copyright © 2011 Korean Diabetes Association http://e-dmj.org
Diabetes Metab J 2011;35:529-535
http://dx.doi.org/10.4093/dmj.2011.35.5.529
pISSN 2233-6079 · eISSN 2233-6087
Comparison of Vildagliptin-Metformin and
Glimepiride-Metformin Treatments in Type 2 Diabetic
Patients
Hyun Jeong Jeon, Tae Keun Oh
Department of Internal Medicine, Chungbuk National University, Cheongju, Korea
Background: The present study investigated the efficacy and safety of vildagliptin-metformin treatment compared to those of
glimepiride-metformin treatment for type 2 diabetes.
Methods: In a randomized, open-label, comparative study, 106 patients with type 2 diabetes were enrolled. The primary end-
point was a reduction in HbA1c from baseline and secondary endpoints included fasting plasma glucose (FPG) or 2-hour post-
prandial glucose (2h-PPG) reduction from baseline, as well as HbA1c responder rate and HbA1c reduction according to baseline
HbA1c category.
Results: Comparable HbA1c reduction was observed with a mean±standard deviation change from baseline to the 32-week
endpoint of -0.94±1.15% in the vildagliptin group and -1.00±1.32% in the glimepiride group. A similar reduction in 2h-PPG
(vildagliptin group 3.53±4.11 mmol/L vs. the glimepiride group 3.72±4.17 mmol/L) was demonstrated, and the decrements in
FPG (vildagliptin group 1.54±2.41 mmol/L vs. glimepiride group 2.16±2.51 mmol/L) were not different between groups. The
proportion of patients who achieved an HbA1c less than 7% at week 32 was 50.1% in the vildagliptin group and 56.0% in the
glimepiride group. An average body weight gain of 2.53±1.21 kg in the glimepiride group was observed in contrast with the
0.23±0.69 kg weight gain noted in the vildagliptin group. A 10-fold lower incidence of hypoglycemia was demonstrated in the
vildagliptin group, in addition to an absence of severe hypoglycemia.
Conclusion: Vildagliptin-metformin treatment provided blood glucose control efficacy comparable to that of glimepiride-met-
formin treatment and resulted in better adverse event profiles with lower risks of hypoglycemia and weight gain.
Keywords: Diabetes mellitus, type 2; Glimepiride; Metformin; Vildagliptin
Corresponding author: Tae Keun Oh
Department of Internal Medicine, Chungbuk National University Hospital,
410 Gaesin-dong, Heungdeok-gu, Cheongju 361-711, Korea
E–mail: tgohkjs@chungbuk.ac.kr
Received: Jan. 11, 2011; Accepted: Jun. 9, 2011
INTRODUCTION
Several clinical studies have demonstrated that tight glycemic
control is necessary to prevent diabetic complications in type
2 as well as type 1 diabetic patients [1-4].
However, tight glycemic control is definitely associated with
frequent hypoglycemic attacks, and several papers recently re-
ported that aggressive glucose control did not produce any
cardiovascular (CV) benefits but rather induced severe hypo-
glycemia and increased mortality in type 2 diabetes mellitus
patients [5-7]. Many type 2 diabetic patients require more
than two oral hypoglycemic agents because treatment with a
single agent often results in sub-optimal outcomes. Combina-
tion therapy early in diabetes management would be a proper
approach considering the complex pathophysiologic defects
associated with diabetes.
Recently, several new classes of oral hypoglycemic agents
have been introduced [8-10]. Vildagliptin is an oral and highly
Original Article

Page 2

530
Jeon HJ, et al.
Diabetes Metab J 2011;35:529-535http://e-dmj.org
selective dipeptidyl peptidase-4 (DPP-4) inhibitor which pre-
vents the rapid degradation of endogenous glucagon-like pep-
tide-1 (GLP-1) and glucose-dependent insulinotropic peptide
(GIP) and increases the levels of the intact, active form of en-
dogenous GLP-1. Vildagliptin improves glycemic control in
type 2 diabetic patients either as a monotherapy or adminis-
tered in combination with metformin, sulfonylurea, thiazoli-
dinedione or insulin [11-14]. Improvements in glycemic con-
trol are mediated primarily by increased insulin secretion and
the suppression of glucagon secretion. Both of these effects
depend on plasma glucose concentration, indicating that insu-
lin secretion is suppressed and glucagon secretion is stimulat-
ed under low-blood glucose conditions. Vildagliptin has not
demonstrated any effects on body weight and does not evoke
severe hypoglycemia.
Sulfonylurea is a potent oral hypoglycemic agent and is
commonly prescribed as a monotherapy or as a component of
combination therapy for the treatment of type 2 diabetic pa-
tients over 60 years of age [15,16]. Although sulfonylurea is
well-known as being effective in lowering blood glucose, it
also induces body weight gain and severe hypoglycemia.
In the present study, the efficacy and safety of vildagliptin-
metformin treatment compared to those of glimepiride-met-
formin treatment were evaluated over 32 weeks in type 2 dia-
betic patients.
METHODS
Subjects
The present study was randomized, open-label and compara-
tive. Patients who attended the outpatient diabetic clinic of
Chungbuk National University Hospital were included, and
exclusion criteria were evaluated during screening (week -2,
visit 1). Type 2 diabetic patients with HbA1c levels greater than
7.0% who were naïve or were receiving monotherapy with oral
hypoglycemic agents such as glimepiride (2 to 4 mg) or met-
formin (500 to 1,000 mg) for less than six months prior to the
visit were eligible to participate in the present study. All the
patients who received previous medications had to undergo a
wash-out period of at least two weeks. Patients with a history
of diabetic ketoacidosis, clinically significant liver or renal dis-
ease, congestive heart failure requiring pharmacological treat-
ment, coronary artery percutaneous intervention or unstable
angina within the past six months, and those over 80 years of
age were excluded from the present study. Any of the follow-
ing laboratory abnormalities at screening also precluded par-
ticipation: serum creatinine >133 μmol/L, alanine amino-
transferase or aspartate aminotransferase >3 times the upper
limit of normal and total bilirubin >34 μmol/L.
All patients provided written informed consent prior to
participation in the present study. The study protocol was re-
viewed and approved by the Institutional Review Board of
Chungbuk National University Hospital. Eligible patients were
assigned randomly at a 1:1 ratio at baseline (day 0) to receive
either 50 mg of vildagliptin twice daily or 2 mg glimepiride
twice daily in addition to 500 mg of metformin twice daily for
32 weeks. During the follow-up period, glimepiride-metfor-
min treatment was down-titrated in cases of recurrent hypo-
glycemia.
Efficacy assessments
Assessments included overnight fasting plasma glucose (FPG),
2-hour postprandial glucose (2-h PPG), HbA1c, fasting insu-
lin, fasting C-peptide, body weight, and vital signs at each
scheduled visit (week 0, 4, 12, 24, and 32). Homeostasis model
assessment of β-cell (HOMA-β) and homeostasis model as-
sessment of insulin resistance (HOMA-IR) were evaluated to
assess pancreatic β-cell function and changes in insulin resis-
tance, respectively. The primary efficacy endpoint was a change
in HbA1c from baseline at week 32. Secondary efficacy end-
points included FPG or 2-h PPG reduction from baseline, as
well as HbA1c responder rate and HbA1c reduction according
to baseline HbA1c category. All adverse events and tolerability
values were monitored throughout the study. Hypoglycemia
was defined as a finger stick glucose concentration of less than
3.9 mmol/L without loss of consciousness. Severe hypoglyce-
mia was defined in the patients with transient dysfunction of
the central nervous system who were unable to treat them-
selves.
Statistical analysis
Vildagliptin was regarded as comparable to glimepiride if the
upper boundary of the two-sided 90% confidence interval
around the between-group difference in HbA1c was less than
the predefined margin, δ=0.3%. The data were expressed as
mean±standard deviation. All statistical tests were performed
using the SPSS version 12.0 software package (SPSS Inc., Chi-
cago, IL, USA). The means and frequencies of variables were
evaluated using Student’s t-test and the χ2 test, respectively. All
P values less than 0.05 were regarded as statistically significant.

Page 3

531
Comparison of vildagliptin and glimepiride
Diabetes Metab J 2011;35:529-535http://e-dmj.org
RESULTS
Clinical characteristics
The present study randomly enrolled 106 patients (52 patients
in glimepiride-metformin group and 54 patients in vildagliptin-
metformin group). One patient in the glimepiride-metformin
group and three patients in the vildagliptin-metformin group
were eliminated from the study in the first four weeks due to
recurrent hypoglycemia or intolerable gastrointestinal (GI)
trouble, respectively. Finally, 51 patients (94.4%) in the vilda-
gliptin-metformin group and 51 patients (98%) in the
glimepiride-metformin group completed 32 weeks of treat-
ment. The patients who discontinued medication due to adverse
events were not included in the final analysis. The final analyzed
patients consisted of 66 males and 36 females, with a mean age
of 54 years and an average duration of diabetes of 5.9 years.
The groups were balanced at baseline with a mean HbA1c
of 8.01 vs. 8.13% and FPG of 8.78 vs. 9.34 mmol/L. Table 1
summarizes the baseline demographic and clinical character-
istics of the patients.
Efficacy of glucose control
Mean FPG±standard deviation was decreased from 8.78±
2.32 mmol/L at baseline to 7.24±1.64 mmol/L at week 32 in
the vildagliptin-metformin group and from 9.34±2.14 mmol/
L to 7.18±1.91 mmol/L in the glimepiride-metformin group.
The reductions in FPG level were comparable between treat-
ment groups (vildagliptin-metformin 1.54±2.41 mmol/L vs.
glimepiride-metformin 2.16±2.51 mmol/L, P=0.508; Fig. 1A).
Both treatments showed similar efficacy with regard to the 2h-
PPG level; the reduction was 3.53±4.11 mmol/L (from 13.67±
3.59 to 10.14±3.46) in the vildagliptin-metformin group and
3.72±4.17 mmol/L (from 14.44±2.22 to 10.72±3.36, P=0.950)
in the glimepiride-metformin group (Fig. 1A).
The mean HbA1c was decreased from 8.01 to 7.07 (1.21%
to 0.81%) in the vildagliptin-metformin group and from 8.13
to 7.13 (0.86% to 0.81%) in the glimepiride-metformin group;
the decrements in HbA1c were also comparable between
groups (vildagliptin-metformin 1.15 (-0.94%) and glimepiri-
de-metformin 1.32 (-1.00%), P=0.855; Fig. 1B).
At week 12, when HbA1c was measured first after the initia-
tion of the study, HbA1c was significantly decreased from
baseline; the decrements in HbA1c did not differ appreciably
between the treatment groups. Additionally, HbA1c remained
stable until the endpoint at week 32 (Fig. 1C).
The proportions of patients who achieved an HbA1c less
Table 1. Demographics and baseline characteristics of patients
Characteristic
Vildagliptin+
Metformin
(n=51)
53.51±10.41
Glimepiride+
Metformin
(n=51)
55.38±10.98 Age, yr
Sex, n (%)
Male
Female
BMI, kg/m2
Duration of diabetes, yr
FPG, mmol/L
2h-PPG, mmol/L
HbA1C, %
Insulin, pmol/L
C-peptide, nmol/L
HOMA-β
HOMA-IR
35 (68.6)
16 (31.4)
22.69±7.75
5.89±1.64
8.78±2.32
13.67±3.59
8.01±1.20
95.21±48.93
1.01±0.70
44.31±20.93
5.11±2.63
31 (60.8)
20 (39.2)
23.07±4.24
5.92±1.74
9.34±2.14
14.44±2.22
8.13±0.86
98.51±79.21
0.99±0.67
41.90±20.93
5.16±6.51
BMI, body mass index, FPG, fasting plasma glucose; 2h-PPG, 2-hour
post-prandial glucose; HOMA-β, Homeostasis model assessment of
β-cell; HOMA-IR, homeostasis model assessment of insulin resis-
tance.
Fig. 1. (A) Mean fasting plasma glucose (FPG) and 2-hour
post-prandial glucose (2h-PPG) reduction in vildagliptin or
glimepiride treatment. (B) Decrement of HbA1c on the week
32 end-point in vildagliptin or glimepiride treatment. (C) Mean
HbA1c changes during 32 weeks in vildagliptin or glimepiride
treatment.
FPG 2h PPG
Vildagliptin Glimepiride
0
-1
-2
-3
-4
-5
Change from baseline
in FPG and 2h-PPG (mmol/L)
Vildagliptin Glimepiride
0
-0.5
-1
-1.5
Change from baseline in
HbA1c (%)
AB
C
HbA1c (%)
8
6
4
2
0


0 12 24 32
week
Vildagliptin
Glimepiride

Page 4

532
Jeon HJ, et al.
Diabetes Metab J 2011;35:529-535
http://e-dmj.org
than 7% at the endpoint were also similar in the vildagliptin-
metformin (50.1%) and glimepiride-metformin (56%) treat-
ment group. In the subgroup analysis of baseline HbA1c cate-
gory, both treatments showed similar HbA1c reduction effi-
ciencies in all categories: greater reductions in HbA1c were
noted in patients with higher baseline HbA1c concentrations
(Fig. 2). The difference between the responders vs. non-re-
sponders was assessed, but no differences were observed be-
tween the two groups, with the exception of glucose and HbA1c
levels.
Baseline fasting insulin, C-peptide, HOMA-β, and HOMA-
IR were similar in the vildagliptin-metformin and glimepiride-
metformin groups. During the treatment period, the changes
in HOMA-β and HOMA-IR were comparable between the
two groups (Table 2).
Changes in body weight
Mean body weight at baseline was 66.2±7.8 kg and 69.7±10.2
kg in the glimepiride-metformin and vildagliptin-metformin
groups, respectively. At week 32, significant weight gain was
observed in the glimepiride-metformin group (2.35±1.21 kg)
compared to that in the vildagliptin-metformin group (0.23±
0.69 kg) (P<0.05).
Tolerability
The overall incidence of hypoglycemia was 10-fold higher in
patients treated with the glimepiride-metformin combination
(ten patients vs. one patient in the vildagliptin-metformin
group, P<0.05; Table 3) even though one patient in the
glimepiride-metformin group was excluded from the study
within the first four weeks of enrollment due to severe hypo-
glycemia. A majority of the patients had hypoglycemia that
did not require medical assistance, although one patient who
received glimepiride-metformin required medical assistance
due to loss of consciousness. No severe hypoglycemia occurred
in any of the patients taking vildagliptin-metformin.
The frequent adverse events in the vildagliptin-metformin
group were GI events such as nausea, vomiting and diarrhea.
Three patients in the vildagliptin-metformin group discontin-
ued taking the medication due to intolerable GI problems
within the first four weeks (Table 3). No meaningful changes
were observed in the hematologic, biochemical or urinary
analysis of vital signs, nor were any significant increases in liv-
er function or muscle enzyme values observed in either group.
Fig. 2. Mean HbA1c reduction according to the baseline
HbA1c category.
Vildagliptin Glimepiride
HbA1c<8 % 8≤HbA1c<9 % HbA1c≥9 %
0
-1
-2
-3
-4
Table 2. Baseline and post-treatment mean insulin, C-peptide, HOMA-β, and HOMA-IR
Variable
Insulin, pmol/L
Vildagliptin+Metformin
Glimepiride+Metformin
C-peptide, nmol/L
Vildagliptin+Metformin
Glimepiride+Metformin
HOMA-β
Vildagliptin+Metformin
Glimepiride+Metformin
HOMA-IR
Vildagliptin+Metformin
Glimepiride+Metformin
0 week12 weeks24 weeks32 weeks
13.27±6.87
13.74±7.15
15.54±9.76
15.96±10.54
16.67±11.24
17.54±12.54
14.02±9.54
14.28±10.54
3.03±2.09
2.99±2.08
3.34±2.04
3.98±2.98
4.04±2.35
3.95±2.99
4.54±2.15
4.15±2.99
44.31±20.93
41.90±20.93
45.49±21.73
45.51±20.84
42.31±20.41
43.01±24.21
47.28±22.36
48.04±23.64
5.11±2.63
5.16±6.51
4.16±2.26
5.01±2.65
4.69±2.99
5.14±2.21
4.53±3.22
4.98±2.21
HOMA-β, Homeostasis model assessment of β-cell; HOMA-IR, homeostasis model assessment of insulin resistance.

Page 5

533
Comparison of vildagliptin and glimepiride
Diabetes Metab J 2011;35:529-535
http://e-dmj.org
DISCUSSION
In the present study, the vildagliptin-metformin treatment
showed an HbA1c reduction comparable to that of the glimepiri-
de-metformin treatment over a 32-week period. The propor-
tions of patients who achieved an HbA1c of less than 7.0% and
HbA1c reduction according to the baseline HbA1c category
were similar between treatments. Furthermore, vildagliptin-
metformin treatment did not evoke weight gain and provided
definite advantages in terms of hypoglycemia incidence reduc-
tion.
Regarding safety, the vildagliptin-metformin treatment had
a favorable hypoglycemic profile: a 10-fold lower incidence of
hypoglycemia, with no severe hypoglycemia observed. Several
papers have recently been published regarding the association
between hypoglycemia and adverse clinical outcome [9,11]. In
the Action to Control Cardiovascular Risk in Diabetes (AC-
CORD) trial, intensive blood glucose control did not produce
any benefits with regard to CV events but did provoke unan-
ticipated excess mortality: 19 of the 41 deaths were attributed
to unexpected CV disease, which may have been related to se-
vere hypoglycemia [5].
In another trial, the Action in Diabetes and Vascular Dis-
ease: Preterax and Diamicron Modified Release Controlled
Evaluation (ADVANCE), severe hypoglycemia was strongly
associated with increased risks of adverse clinical outcomes.
During the follow-up period, severe hypoglycemia was associ-
ated with a significant increase in the adjusted risks of major
macrovascular events (hazard ratio [HR], 2.88), major micro-
vascular events (HR, 1.81), death from a CV cause (HR, 2.68),
and death from any cause (HR, 2.69; P<0.001 for all compari-
sons) [17]. The results of these analyses showed that the previ-
ous severe hypoglycemia occurrence was likely to constitute a
marker of vulnerability to a primary CV event or death re-
gardless of the treatment arm. Considering the potential CV
risk associated with severe hypoglycemia, the use of medica-
tions associated with a low risk of hypoglycemia for the con-
trol and management of blood glucose in patients with type 2
diabetes would be prudent. Three patients in the vildagliptin-
metformin treatment group discontinued taking medication
in the first four weeks due to intolerable GI side effects, but
these effects were likely attributed to metformin rather than
vildagliptin since the GI side effects in the patients disappeared
after vildagliptin monotherapy. Several patients in both treat-
ments complained of GI side effects such as metallic taste, re-
duced appetite, or dyspepsia. However, the majority of GI
symptoms disappeared over time in both treatment groups,
and the patients who complained of GI troubles in the initial
period of treatment eventually grew tolerant of the medica-
tions. DPP-4 inhibitors increase endogenous GLP-1 levels via
inhibition of the DPP-4 enzyme. Increased GLP-1 can physio-
logically inhibit the satiety center in the brain, reduce appetite,
and delay stomach emptying [18]. Vildagliptin-metformin
treatment has a greater potential to evoke GI side effects com-
pared to glimepiride-metformin treatment in terms of the rel-
evant pharmacological mechanisms, although previous stud-
ies did not show any further increase in GI side effects between
metformin monotherapy and metformin-vildagliptin combi-
nation therapy [19,20]. Although there is no current obvious
explanation why only patients in the vildagliptin-metformin
treatment group experienced intolerable GI trouble, initiating
vildagliptin treatment with a low-dose of metformin in order
to ameliorate the GI side effects is advisable.
In the present study, changes in body weight differed be-
tween the treatment groups. At 32 weeks, body weight was
unchanged in the vildagliptin-metformin treatment group,
whereas the patients treated with glimepiride-metformin evi-
denced weight increase (2.35 kg) relative to baseline. The re-
sults were consistent with findings reported in other studies
involving vildagliptin or sitagliptin in combination with met-
formin [21,22]. In another study comparing the efficacies of
Table 3. Safety summary
Variable
Vildagliptin
(n=51)
5
0
3
1
4
Glimepiride
(n=51)
10 One or more AE
SAEs
Discontinuations due to AE
Hypoglycemia (patient number)
Overall gastrointestinal AE
Selected gastrointestinal AE
Abdominal pain
Nausea
Vomiting
Diarrhea
1
1
10
0
0
1
2
1
0
0
0
0
The patients who discontinued medication due to AE were not con-
tained in final analysis.
AE, adverse event; SAE, serious adverse event, which means to need
medical assistance.